[peterpan188]

Hi there, my friend's bday is comming up in around 2 months, and we, as good buddies, decided to make him a good MAXED out Millet Hybrid, in order to hurt his future wallet futher. He does own some headphones, including HD600, but was never fond of the idea of amps. The reason MH was chosen is becasue 1) I have a MH board and a DB board left 2)It uses tube, mainly enhance the look, 3)It's within our bugdet (max. $500). So now here are some questions about part selections and general opinions. Any input will be much appreciated, and I actually would like to gather some knowledge about choosing top components, so others that would like to do the same can have a look at this thread.

1) It is still undecided whether the STEPS and MH should be in the same enclosure. The advantage of separate chassis would be easier to transport (we all still live in college dorms) and the STEPS can be reused for other amps that he/we might eventually build. Combined chassis would give more professional look, easier cable management and probably nicer casing. What is your insight here?

2) I looked into the BOM of STEPS, and I did some searches in old thread, and I couldn't really find anyone who really made a big change in parts for the STEPS, including caps (well, in most cases, that's where we make changes). Is it that cost of part subsituition isn't worth the change in sound quality? Or is it the suggested Panasonic filter cap and Kemet interference suppressor caps are good enough as concerned?

3) This is also related to choice between casing, for a separate STEPS, a LM338T with Ameco 70062 would be the choice as amb suggested, so then it will work with other higher current amps. Would it be a waste to just use this combo if its just for MH? becuase I usually populate the board and then do the casing, and desoldering the IC with the heatsink would be a pain.

4) I have been digging out threads, such as this one , and I do understand that the MH has its own limitation, ie. distortion, so being insane about maxxing it out would not be worth the price.

I am thinking of using all BG NX caps, and I have read from n_maher that there is a "Super-E" Configuration that uses 1 or more BG caps as bypass. And I have also read, forgive me for forgetting who said it, that BG don't need any bypass film caps in the signal path. What's your opinion on these idea? How did the sound turn out? I would still like to keep the board looking neat, would Super-E be too many caps sticking around?

For resistor, I am thinking if I should go all out and use boutique resistors, such as Riken, PRP, Holco and Kiwame, or selectively use them where everywhere else would be RN60. How do these boutique resistors compare?

Tubes, I guess I will find matched NOS pairs for all 3 tubes. Does the brand of the tubes really matter? Where would I be able to source them? I got myself a Tung-Sol before and it sounded fine but I never did any comparsion.

I will start off with stacked BUF634, and after I make sure everything's working, I will populate a diamond buffer in, but that's the next thing for now.
Jacks would mostly be Cardas female RCA and Neutrik Locking 1/4", I would like to include loop-out, but not sure how to do it, can anyone help here?

5) This is more general. If I were to build a separate STEPS, I would like to have a feature that 1 Power LED will light on when I turn the switch on, and 1 Indicator LED will light on when I plug a DC plug into the output DC jack. How can I do that? I have been using Kobiconn switched DC jack, and I think it is normally closed, how can I use that to build the feature?

Thanks, guys, lots of questions, and any input will be much apppreciated. Thanks again for n_maher for the work on MH, that eventually led to this many questions.

Peter

 

[NeilR]

I built mine in separate 1455N1601 cases. In separate cases there is no chance for any interference from the transformer. It is measurable, probably not really audible. Separate cases are probably easier to work with (and probably cheaper than a single large higher end case) and work better with limited desk top space. If you build more amps you have a PSU pre-built. I use my Steps with 3 different amps (Millet/PPA/M3). I like that because I can build an amp and then decide on a PSU later, or never build another PSU, saving $100.

I actually think a separate PSU is "more professional" from certain points of view. An integrated case is just bigger and maybe more visually impressive, for better or worse.

Aesthetically, I think Millets look better if there is some symmetry to the layout in terms of the tubes, which usually poke out of the top. The Hammond case gives you a perfectly symmetrical layout, where an integrated case will give you a lopsided look in one direction or the other- you have to chose which looks better, I guess.

If you are interested in those last 0.1% tweaks that BGs might give you, consider that the most important thing is probably to minimize the input wiring lead lengths, including any leads to the POT. A board mounted pot will have the shortest (zero) additional lead length and the least chance of picking up interference or noise. If you build an integrated case, your layout considerations may lead you to panel mount the POT away from the board. The input jack wiring will also be shortest in the Hammond case.

If you use the matching Hammond cases, and you screw up your panels, it is possible to source spare panels to try again, although they have to be ordered in lots of 10 and special ordered. If you have not done a lot of case work that might be a consideration. Or maybe redo the panels in acrylic, which I am thinking about.

I used an LM338 in my Steps, based on AMB's discussions related to M3's. That is a very tough little regulator and a good $2 upgrade.

It's all personal choice, really; you'll get opinions every which way, especially with the upgrade parts you mention. I have somehow managed to get listenable music out of my amps without using BG's or $5 resistors

Here is my Millet, with matching Steps in a Hammond.

 

[tomb]

Very nice work, Neil.

I see you put remote leads for the bias test points in back. Do you leave the top open? I've wondered about the remote bias points, but have opted for a couple of keystone clips, instead (not on the adjacent ground points - that will be somewhere else). It seemed a false savings unless you remotely located the pot trimmers, too ... unless you just leave the top open. On the other hand, I've seen some people drill holes in the top for trimmer adjustment - that seems dangerous, too - unless the top was transparent.

You also used the same size case for the STEPS - did that give you needed extra room for the power socket, etc.? It looks so. I would assume that you keep the top on the STEPS.

Last question - is that a parallel-wired, optional 3.5mm input jack in front? I guess with that you can use patch cables for stationary audio equipment in back, then the 3.5mm input jack in front gives you an iPod/PCDP option. Following your advice with short leads, does this cause a problem with the input leads along the whole length of the Millett?

 

[n_maher]

Quote:

Originally Posted by peterpan188 1) It is still undecided whether the STEPS and MH should be in the same enclosure... What is your insight here?

Separate will probably be cheaper. After that I think it's pretty much personal preference.
Quote:

2) I looked into the BOM of STEPS...

I wouldn't feel the need to step outside the normal STEPS parts list if I were you.
Quote:

4) I am thinking of using all BG NX caps, and I have read from n_maher that there is a "Super-E" Configuration that uses 1 or more BG caps as bypass.

Can't say I recommend using the NX caps, no one is sure that it'll work yet and it's a pretty big gamble $$ wise.

Quote:

For resistor, I am thinking if I should go all out and use boutique resistors, such as Riken, PRP, Holco and Kiwame, or selectively use them where everywhere else would be RN60. How do these boutique resistors compare?

The resistors are short money any way you slice it, I'll be using a mix of all of the above most likely.

Quote:

Tubes, I guess I will find matched NOS pairs for all 3 tubes. Does the brand of the tubes really matter? Where would I be able to source them? I got myself a Tung-Sol before and it sounded fine but I never did any comparsion.

Can't say I'm going to take the time to compare them

, like you said the amp has its limits.
Quote:

Thanks, guys, lots of questions, and any input will be much apppreciated. Thanks again for n_maher for the work on MH, that eventually led to this many questions.

Thanks, but there's a whole list of folks who were involved in that effort that deserve just as much recognition.

Best of luck with the build.

Nate

 

[NeilR]

Good observations and questions, tomb!

Very nice work, Neil.

Thanks!

Quote:

Originally Posted by tomb I see you put remote leads for the bias test points in back. Do you leave the top open? I've wondered about the remote bias points, but have opted for a couple of keystone clips, instead (not on the adjacent ground points - that will be somewhere else). It seemed a false savings unless you remotely located the pot trimmers, too ... unless you just leave the top open. On the other hand, I've seen some people drill holes in the top for trimmer adjustment - that seems dangerous, too - unless the top was transparent.

When I am in a playing around mood, I do leave the top off. Usually it is in place. I would like to put clear tops with holes drilled to access the trimmers. As you suggest, I don't care to poke around blindly with the stock metal top. As strange as this may sound, I was prompted to put the test sockets in the back because on a number of occasions, I biased the amp, pulled the tubes, replaced the cover, and then was unsure if I had switched tubes! By socketing the rear, I can at least verify that the tubes were not switched and preiodically check for bias drift. You are correct that it needs a clear top with holes drilled for the trimmers to make it work right. I considered panel mounting pots but space is limited in the back and I did not want more multiple inches of wire running around inside.

The one mistake I made is that the L and R sockets are too close together to use simultaneously. At the time I did not have a 2nd meter (I now have 4). Even so, I thought about it when I drilled the panel but I could not come up with an alternate arrangement I was comfortable with. The caps and other parts sitting near the back end of the board make it difficult to add things to the rear panel... I ran into all sorts of clearance issues. When I am rebiasing or probing around, I hook a grabber lead to the top socket's big solder lug and put a meter probe into the bottom socket, giving me simultaneous monitoring, so in the end it worked out ok. I hook the meter grounds to the bare wire connecting the input jack grounds.

Quote:

Originally Posted by tomb You also used the same size case for the STEPS - did that give you needed extra room for the power socket, etc.? It looks so. I would assume that you keep the top on the STEPS.

For two reasons: 1) I wanted a matching case size for aesthetics and 2) I was concerned about the power inlet wiring. I used stiff wire and it is a tight fit down to the pads. If I wanted the smaller case, though, I would have just done that. On the front, it does allow a little extra working room for the switch lugs.

That Steps was my first metal casing job and a very early project. If I were doing it again, I would be less concerned about tight quarters. I have a 2nd Steps in the works, but it is going in the same size case just for aesthetics. I also like to be able to stack the PSU and the Amp where table top space is tight and I like the PSU on the bottom for some reason, although it makes a little more sense to stack it on top (for better ventilation).

Quote:

Originally Posted by tomb Last question - is that a parallel-wired, optional 3.5mm input jack in front? I guess with that you can use patch cables for stationary audio equipment in back, then the 3.5mm input jack in front gives you an iPod/PCDP option. Following your advice with short leads, does this cause a problem with the input leads along the whole length of the Millett?

That is a parallelled OUTPUT jack. For good reasons that have not survived the test of time and my DIY redirection, I have a (very expensive) Cardas cable for my HD-650's terminated in an 1/8" plug so I figured that would be the most robust way to connect them. I was concerned about the run to the back of the board, but this was done after I put the 1/4" jack in and I didn't want to bother rewiring that jack to do it "right". I have done a lot of RMAA testing and I do not think the extra wire run adds any noise or interference. I built a couple of nice 1/8 ---> RCA ICs for my iPod connection and I prefer a rear connection for the inputs. I did not bother to do this on my other desk top amps because I have a DIY adapter that is more or less working ok and eventually I plan to replace that Cardas cable with a proper 1/4" plugged version. I was waiting for Cardas to release those connectors but that idea seems to have died.

I don't like the look of that CUI 1/8" jack because of the chrome. I recently used the isolated jacks that Tangent listed on the Pint (rear entry with a black snout with gold trim). That would look much better, is cheaper, and I can't tell the difference quality-wise. I think that is a nice jack. However, I think it takes a smaller panel hole so I can't easily swap jacks. I might also replace the 1/8" jack with a power switch. Decisions, decisions.....

Regards,
Neil

 

[NeilR]

Tomb,

I could not get the Keystone clips to fit in the bias pads without cutting off a prong. That was my first attempt. I have always been leery of the close proximity of the bias + to the bias gnd pad. I blew up a couple Intersil buffers after various probe slips- those buffers (since retired for the dDB boads) can't survive a short at the bias pads- not even the slightest fraction of a second. One reason for the panel mounted probes was just because I got tired of blowing up those $5 chips.

Neil

 

[NeilR]

P.S. One other benefit of the longer Steps case is that it allowed me to route the power switch and DC out wires under the board for a neater build. I got that idea from studying AMB's build images on his site. He has a knack for build aethetics (and quality). If I had a need for the shorter case I would just run the wires above deck but it was an added minor benefit...

The Steps board was optimized for a power switch on the rear and DC out on the front. I just had to be different. It's a personality defect.

Neil

 

[tomb]

Neil,
Thanks for the answers to my questions. As usual, you made everything quite clear. I will have to try something creative, I guess - about the bias points. I have four boards and was going to try all four options of buffers at some point - stacked BUF634's, the OPA551, the Intersil's, and the DB's. I already have all but the DB's and Nate is going to help us with that soon.

Do you have a part number handy for those test probe sockets?

 

[sejarzo]

Quote:

Originally Posted by peterpan188 Tubes, I guess I will find matched NOS pairs for all 3 tubes. Does the brand of the tubes really matter? Where would I be able to source them? I got myself a Tung-Sol before and it sounded fine but I never did any comparsion.

Maybe you caught some of my previous posts on this topic, but every 12FK6 in my possession--RCA, TungSol, DuMont--has the 12FK6 designation surrounded by the octagon (in addition to the TungSol or DuMont markings) that indicates the tubes were manufactured by RCA and simply rebranded in the case of TungSol or DuMont.

I'd be really interested to hear from anyone that has any 12FK6's that don't have the octagon either alone, or in addition to the markings of a non-RCA brand.

And as to matching, one supplier told me that the only thing that can be checked on these tubes is transconductance, and he said that he could randomly pick any two new 12FK6's from his stock and they would qualify as "matched" as there was just so little difference between them.

I've had back luck with a couple of pairs of 12AE6 or 12AE6A's in my MH.....either they were microphonic, or seemed to distort on percussion-heavy passages when listening to classical music at normal volumes.

Any unused compatible tube for the Millett is "NOS" as those types haven't been manufactured since the early 60's. And it seems as if the supply is running low as there was no demand for these types after the demise of tube AM radios in vehicles--many were simply discarded years ago when thought to be not worth keeping in stock!

 

[NeilR]

Quote:

Originally Posted by tomb Neil, Do you have a part number handy for those test probe sockets?

Tip Jack - BlackMouser530-105-1043-1
Tip Jack - RedMouser530-105-1042-1
Tip Jack - WhiteMouser530-105-1041-1

One outside the box idea would be to mount the trimmers to the bottom of the board. You would then have to raise the board high enough to accomodate that, which would be a problem in the standard Hammond. I was thinking about something like that for a Tread supply that I want to use as a poor man's bench supply, although in that case the pot would go off board.

Good luck with the DB's. They are fun to build. Be sure to use the SIP sockets for the test points. When you get it cased up, run some temporary wires from the SIP sockets up through the tube holes. Take a look at what happens to the bias when the oven door is shut

Regards,
Neil

 

[tomb]

Quote:

Originally Posted by sejarzo Maybe you caught some of my previous posts on this topic, but every 12FK6 in my possession--RCA, TungSol, DuMont--has the 12FK6 designation surrounded by the octagon (in addition to the TungSol or DuMont markings) that indicates the tubes were manufactured by RCA and simply rebranded in the case of TungSol or DuMont. I'd be really interested to hear from anyone that has any 12FK6's that don't have the octagon either alone, or in addition to the markings of a non-RCA brand. And as to matching, one supplier told me that the only thing that can be checked on these tubes is transconductance, and he said that he could randomly pick any two new 12FK6's from his stock and they would qualify as "matched" as there was just so little difference between them. I've had back luck with a couple of pairs of 12AE6 or 12AE6A's in my MH.....either they were microphonic, or seemed to distort on percussion-heavy passages when listening to classical music at normal volumes. Any unused compatible tube for the Millett is "NOS" as those types haven't been manufactured since the early 60's. And it seems as if the supply is running low as there was no demand for these types after the demise of tube AM radios in vehicles--many were simply discarded years ago when thought to be not worth keeping in stock!

Ummm ... I have a pair of GE 12FK6's with no octagon anywhere on the tube. Most likely GE was more than capable of making all of their own tubes.

My Tung-Sol's indeed have the octagon - as you say.

 

[hell0un1verse]

Quote:

Originally Posted by n_maher Separate will probably be cheaper. After that I think it's pretty much personal preference.

Nate, I saw the 16x12x3 Par-Metal chassis case you used for an M^3 on headwize.com M^3 build discussion thread. Do you think electrical interference wise(in measurement) this big case is almost as good as a separate cases approach? I am also thinking about which way to go for my newly built M^3 and STEPS. Aesthetically, a single big case looks quite nice and attractive, but separate cases look more high end, high end HIFI stuff always tend to separate different function units in separate cases.

Thanks.

 

[peterpan188]

Sorry for leaving this thread alone for so long, because I went to a little vacation, got my cellphone snatched in the subway, and never had to chance to go on Head-Fi.

It's a very nice job you have done with the MH there, Neil. I was also thinking of putting remote bias point test lead, after seeing what steinchan did. I like the way he design it that there is no need to open up the case to adjust the trimpot. I thought about the casing, got a conclusion that I should do the STEPS and MH separately. I have a big piece of 1/8" thick plexiglas in hand, so I might just use that do the top cover. I thought about the input wire length. I personally prefer the volume cotrol being off center toward the right, so I guess I will have to case mount the POT, but I will keep it in mind to minimize the length.

As I was working on the BOM with excel, I found that the transformer 70062 that I was intending to use is out of stock with no ETA. So I will 70063, which is rated for 15VAC@1.664A. Anyone find that a problem?

Regulator will be a LM338T. Caps are Panasonic ECG suppressor , Kemet bypass, BC Filter and Nichicon UPW. I will be using DC locking jack from Kobiconn, and built my own DC cable using 18awg wires. Total comes out to be around $70 without case for the STEPS

For MH, I think I will use Blackgate NX 680uF for C7 , and Std 470uF for C1, C9, C10, and Std 220uF for C2, since they are all available from partsConnexion. I have read that for C4, it would be better to use Panasonic ECQ-P, but on the BOM on DIYforum listed a ECG, what's the difference? Also, for other bypass caps, the BOM listed a Wima MKP10 0.1uF 10%, but I found the exact 0.22uF but with 20%, both on Mouser, which one should I use? I have also read that I should not add bypass caps with Blackgates for the optimal perforance, is that true?

Resistors will be all PRP 1/2W, except R4 the output 22ohms, which will be Riken.

I will be installing all 4 sockets on the board for buffer, because I will be using BUF634 for inital setup and diamond buffer after everything works out.

Input wil be a pair of Cardas GFRA RCA, output will be Neutrik locking headphone jack. I would also like to add loopout and preamp out, with my understanding, loopout is just simply connecting input RCAs and loopout RCAs together, and preamp out is just a parallel output along with the headphone out, is that right?

As for the tubes, where did you guy get your brand name tubes, such as RCA, Tungsol, etc? I looked into tubedepot and the other 2 stores listed in DIYforum, but they don't offer me opinions to choose brand. I understand that it wouldn't actually affect the sound, but I would like to be a good logo on it

. Last time I order from tube depot, they sent me a pair of 12AE6A with nothing on it.

Thank you for all your input.

 

[NeilR]

peterpan,

Sorry to hear about your cell phone. At least they didn't get your portable amp and cans, if you have them

I would like to redo my top in acrylic for the same reason. It would make the tube rolling much easier.

The standard advice here is NOT to use a 30V transformer because you are dropping at least 6 additional volts across the regulator "unnecessarily". You would be better off using the unpotted 24V (TE62062-ND) version, which is available. According to Tangents PS configurator, a 30V transformer is acceptable for the 350ma draw that you will get with the discrete boards, but it will run fairly hot, but apparently not in his "warning zone". My Step's reg runs "hot" with the 24V part. At least that is what my finger tells me, regardless of the actual tempeorature and specs.

The capacitor Connoisseurs here can correct me if I'm wrong but I think Panasonic "ECG" and "ECQ-P" is the same part. The manufacturer part# starts with "ECQ-P..." and the line is called ECG by Digikey, but not by Panasonic. I don't have a board in front of me at the moment, but any time you use those bigger ECQ-P caps like the .22uF, you should measure twice and buy once, especially if you are pushing the size of the surrounding parts (thinking of the BGs here). I've had problems fitting that part in a couple of projects although I never tried with the Millet.

Historically, C4 and C7 have been discussed as being in the signal path. C2 and C3 are also in the signal path, and that has been discussed occasionally in the various fora. If you want to max out the amp, you should give C2/C3 the same treatment as C4/C7, although C2 may present some real estate issues.

I used the .1uF Wimas in mine and nothing blew up

. The values are not that critical and some people insist on not using those smaller bypass caps at all. Just to say that you may get more opinions than there are different possible part values. The one that sounds the best will probably be the one you are most comfortable with, regardless of the validity of your reasoning

If it were me, I would leave the bypass caps out and see how it sounds, then put them in and see if you hear a difference. That is a fiuture project of mine that I haven't gotten around to. The fun of building a Millet is that it is cheap (before you add jewelry to taste) and easy to modify and tinker with.

Regards,
Neil




Try setting up some ebay searches for the tubes you want. I got a nice set of 5 Tung-Sol 12FK6's recently.

Regards,
Neil

 

[peterpan188]

Quote:

Sorry to hear about your cell phone. At least they didn't get your portable amp and cans, if you have them

Well, I really can't understand how the majority of the people think these days. I chased the guy down 5 blocks, on BOARDWAY!! NYC!!!, not a single person helped. I know that the general recommandation is not to chase down the person who rob you, since he might have weapons, but come on, theres more than a thousand people on the street, if somebody just gave me a little hand, he would have been caught.

Quote:

The standard advice here is NOT to use a 30V transformer because you are dropping at least 6 additional volts across the regulator "unnecessarily". You would be better off using the unpotted version, which is available. According to Tangents PS configurator, a 30V transformer is acceptable for the 350ma draw that you will get with the discrete boards, but it will run fairly hot.

The reason I want to use the 25VA transformer is that I am not sure whether or not the friend that I am making this for would be interested in some other design, such as the M^3, for the future. I want to build him something that he can reuse, and I saw on amb's site that he used the 70062 on his STEPS to power both MH and M^3. Since 70062 (12V@2A) is out of stock, I chose 70063(15V@1.6A). Not sure if it is necessary. For the concern of heat, I can probably open up a hole right above the heatsink for better thermal distribution.

Quote:

The capacitor Connoisseurs here can correct me if I'm wrong but I think Panasonic "ECG" and "ECQ-P" is the same part. The manufacturer part# starts with "ECQ-P..." and the line is called ECG by Digikey, but not by Panasonic.

I looked into the catelog of Digikey, I found that they listed the ECQ-P on their catelog, but when you enter the part number in the website, they show you its ECG. So I think you are right here. Turns out that what the BOM listed (P3224-ND)and what I found (P3929-ND) are the same capacitor, I think Digikey updated the catelog.

Quote:

Historically, C4 and C7 have been discussed as being in the signal path. C2 and C3 are also in the signal path, and that has been discussed occasionally in the various fora. If you want to max out the amp, you should give C2/C3 the same treatment as C4/C7, although C2 may present some real estate issues.

Exactly what you said. C2 has a smaller diameter, and the best thing I found to fit in would be Blackgate STD 220uF and Cerafine 220uF. I want to order from the same place, from partsConnexion, so Blackgate STD was the pick.

Quote:

I used the .1uF Wimas in mine and nothing blew up . The values are not that critical and some people insist on not using those smaller bypass caps at all. Just to say that you may get more opinions than there are different possible part values. The one that sounds the best will probably be the one you are most comfortable with, regardless of the validity of your reasoning

I was once thinking about going all out and use some Solen, Cardas or whatever crazy film cap there is, but I was think that will probably not as good as leaving them out eventually, especially with Blackgate. I guess I will get the Wima, populate them on the board, and see what comes out if I take them off.

Thanks Neil, I will start some search on eBay.

Thanks,
Peter