[runeight]

Yaaaa!!! I am very happy to get this news.

I assume that you mean TB+L minus TPL is about 13V? And the same for the right channel? This means that the individual triode sections area bit more out of balance than typical, but not critically so. Let it run for a while and see if things change.

What have you set the B+ to be at this point? Whatever it is you can be free to adjust it from it's lowest to highest value using the trimpot.

 

[Uncle Bob]

Yep, TB+L minus TPL is around 13V. TB+L/R is 70V, heater voltage for the tubes is set to 12.6V

Just off to spend some time with my new toy, will report back any observations tomorrow

 

[Ferrari]

Very good news Uncle Bob. You have beaten Gewa, as far as I can see.
Congratulations and enjoy your new toy!

 

[Uncle Bob]

I can't believe I've beaten Gewa, I reckon he's spending too much time listening to it to let us know he's finished. In any case, I still have the casework to do

I think I'll go for one of those Galaxy cases from Italy.

Anyway back to the matter in hand. I bought a 10 pack of tubes so I've gone through them with B+ @60V and get a spread of results from a low of 39.6V to a high of 46. Swapping tubes around doesn't seem to make any difference to the results, the votage readings follow the individual tube. Equally, setting B+ @80V just pushes the result of a given tube up exactly 20V although some of the tubes read 21 or 22 volts higher. I'm not using a branded meter so there's obviously room for error here as well. The max B+ I can get out of my Soha is 99.8V. I've ended up today with two of the closest matching tube pairs reading 63V each with B+@80V and that is the setup for tonight's session.

It was a joy listening to it last night

Tried with my Grado SR225 and liked it a lot. I played Natalie Cole's "A Woman Who Knows" and was particularly impressed on how metallic the struck and brushed cymbals sounded, just like they should do. Her vocals were presented as intimate and warm but quite natural. I then moved onto Dire Straits "Love Over Gold" (yeah I know but it is one of my reference pieces and still one of my all time favourites) and it performed pretty well here except for a slight sense of squashed dynamics if I was being ultra critical. This may be down to the tubes of course, but atm I don't have any other brands/types to compare against - a situation I will have to address quite soon

I then switched to my Senn HD650's and Alan Parsons "I Robot". My initial impression is that I don't like this combo as much as the Grado. Some people refer to the HD650's as dark sounding and that's what they sounded like here. Again a different tube choice might change the sound signature entirely, so ymmv as they say. In any case, some more listening time is needed before any conclusions can be reached for me.

 

[Uncle Bob]

I was thinking last night about casing up my SOHAII and I thought it would be nice to have a bi-color led wired into the E12 circuit so that the led would initially be red to indicate power on and then green when the E12 enables the relay.

Would this be easy to implement?

 

[runeight]

Since you want to use a common cathode LED, this is about the easiest way that I can think of. The PNP devices can be 2N3906, BC560, 2N5087, etc, but anything with hfe over 100 should be suitable.

Set the RLED for 10mA or less. This is to protect the TO92 regulator on the SOHA II board by keeping its power dissipation as low as possible. 10mA should light most LEDs that you might want to use.

 

[bmwpowere36m3]

Quote:

Originally Posted by runeight /img/forum/go_quote.gif Congrats bmw. Looking foward to your case up now and your listening impressions. I assume the build was straightforward with no problems? ....

Thanks, so far I'm really impressed by this amp. Definitely a lot better than the "SS" Hybrid and CMoY it has replaced. I've been listening to it with Senn HD580's and EH 12AU7 tubes thus far and I really enjoy the sound. The amp has been burning in for over now ~90hrs straight (have been listening here and there; at night letting it just run with a "test" set of headphones). I'll say the EH tubes have definitely settled in since when I first listened to them.

The build was straightforward and nothing really stood out as problematic. I only had trouble since I didn't install the pot before testing and was worried something had been damaged. WHen that was alleviated, I finished up the build no problem. Since I changed up my build, I had some "problems" getting everything to fit... eventually figured out the layout.

The EH tubes at first were a little harsh and very in your face, forward soundstage. As they burned in they've settled back, but still have a forward presentation. They have clear, defined highs, good midrange punch, and bass detail. I'd say the only thing they lack is maybe some more bass slam and midrange bloom, however like Holland has mentioned elsewhere they sound wonderful paired with "Metal".

I've listened to music ranging from Pink Floyd, Led Zeppelin to Metallica, Alice in Chains, In Flames, AC/DC, to some classical. It all sounds good and much more detailed than any other amp I've had thus far... when I start rolling tubes I'll be looking for something with a "warmer" sound (for Floyd, Classical, etc..), as the EH's are fairly neutral compared to my "SS" Hybrid which was very "tubey" sounding and a little overly warm.

I'm impressed

and here are some pictures of it cased up for now, it'll stay like this till I get some more time to finish it up with a top. Also the knob is temporary, scavenged from my "SS"...

 

[bmwpowere36m3]

Guess I still have a couple questions

...

First, is the BOM switch supposed to be lit when "OFF" and not lit when "ON". Since I've tried different wiring schemes and it either works that way or not at all.

Second, what's the best way to "reset" or adjust the O/P quiescent current, with the tubes removed, tubes in, doesn't matter? Any advantages to running a higher current?

Third, adjusting the "tail current", again are there any advantages to playing around with this? If I understand correctly the "tail current" (plate current?), plate voltage, and grid voltage sort of go hand in hand. With the Stock figures, plate voltage ~40V, plate current 2ma, the grid voltage is ~-1V (based on EH 12AU7 datasheet). I'm still learning these things, I know you want to keep the grid negatively biased, so if you input from a "CD-source" then you get ~1Vrms. So is -1V grid voltage enough or it just works, I also understand you don't want to much negative grid voltage, sort of counter productive.

So to achieve a -1V grid voltage at ~80V plate voltage, you'd need ~5-6ma plate current... Maybe this is just a fruitless quest, does it really matter I guess. Any helpful advice or source of info to read would be great. Thanks and Happy New Year to all.

 

[runeight]

Nice job on packaging up that build bmw. I was wondering how you were going to fit the heatsink for the heater reg. But you went ahead and attached to the box. Good idea.

The switch is supposed to be lit when ON. This probably means that the power connections to the switch are reversed? Which would mean that my wiring diagram on the website is incorrect. I haven't wired the light because I don't want to have the extra illumination and so haven't verified the wiring.

You can adjust the O/P current anytime, tubes or no tubes, but I recommend removing the headphones first.

You might hear differences when you change the tail current because the idle current is so low. But, to keep the tube stage working properly you'll have to adjust R4 to drop 18V for whatever value of current you choose.

Right now the tail current is 2mA making each triode pass 1mA. !8V, obviously, means 18k at 1mA. If you adjust the tail current to 4mA you'll have 2mA in each triode and you'll need 8k5 (or close) to get the right plate voltage.

Alternatively you can use the 18V zener as shown on the website. Then the plate voltage will be independent of tail current. You may, however, hear zener noise.

You don't really need a grid bias equal to the maximum peak voltage of your input signal. This is because the common cathode input stage has some amount of negative feedback applied. You can probably get away with as low as 300mV before you'll start having problems.

The data sheets for tubes are notoriously bad at these low voltage low current conditions. Often the curves were extended from the actual data points using French curves. Having said that you can try to use the curves to figure out where the bias will be. It will be different for each different tube type and different for tubes of the same type. Don't forget that you can increase the B+ to over 90V. This will help to increase the cathode bias voltage for a given current.

If you don't want to look at the plate curves then just try it and measure the cathode voltage. If it is above 300mV you're probably ok. If not, you'll hear it pretty quickly.

Tube rolling - I have noticed that the 6922/6dj8/8416 tubes have a more tubey sound and better bass compared to the 12au7s types. Also, with your super heater supply you can try those 6n1ps and 12bh7s. I have no idea what they sound like.

 

[runeight]

bmw I have posted a new inlet wiring diagram on the website. I'm pretty sure that if you wire it this way the light will work properly.

 

[Uncle Bob]

Quote:

Originally Posted by runeight /img/forum/go_quote.gif Since you want to use a common cathode LED, this is about the easiest way that I can think of. The PNP devices can be 2N3906, BC560, 2N5087, etc, but anything with hfe over 100 should be suitable. Set the RLED for 10mA or less. This is to protect the TO92 regulator on the SOHA II board by keeping its power dissipation as low as possible. 10mA should light most LEDs that you might want to use.

Thank you for taking the time to work that out for me, it even looks simple enough for someone like me to implement

I'll let you know how I get on.

 

[runeight]

You're welcome.

Here's an alternate way to wire the LEDs that might be preferable because it keeps the LED current out of the ground flow. RLED is calculated using 24V instead of 12V.

Edit: I've changed the drawing slightly to be consistent with common representations of dual LEDs. And, of course, one can always use individual LEDS.

I've also changed one resistor value. It's not important, but it does serve to equalize the base currents for the alternate version.

 

[holland]

That's a really cool idea, Uncle Bob.

 

[DrkRipper]

Alright...question time:

I have everything assembled and voltages set correctly. With tubes in and music playing I only get music from the right channel. TB+L is ~60V and TB+R is ~25V. When the tubes are out they are both 60V. Sometimes when I change the volume the relay trips, the music stops...then a few seconds later it comes back on. Also I may have screwed up a transistor...I did not use a heatsink which I read noobs like me are supposed to use...so it might be a bad transistor. Any ideas?

 

[runeight]

That's good, mostly.

The first thing to do is to replace Q1P. This is the cap multiplier for the right channel. This type of problem is usually caused by a shorted Q1P (or Q2P on the other channel).

Let us know if this works. If not, the next step is to replace Q1 and Q2 in the amp section.

If this doesn't work then we've got some thinking to do . . .

It is normal for fast rotations of the volume control to trip the e12. But if this happens at all times then it also indicates a problem

Were you able to set the bias currents for both the O/P stage and the tube stage?