[Epicurean]

I took your advice n_maher and replaced the output resistors, hooked up my cheap phillips sbc-hp800 and powered up. It's amazing, but it works! At first there was only hiss in the right channel but since i have everything hooked up with alligator clips i rechecked them and the right channel was loose at one point. I haven't tried it with a good source and phones yet though, so I'm not sure if the output trannies are completely intact.

 

[NeilR]

I recently built an M3 and I discovered what was apparently a poor solder joint by running RMAA tests on the amp. The distortion on one channel was high- about 0.5%, where it should have been less than 0.01%. I would never have found that without testing. All the joints looked ok and I only fixed that issue by reflowing all the joints in the affected channel. If your trannies are impaired but basically functional, you may be able to find it by testing it with RMAA.

If you have a good soundcard that can do a loopback, RMAA software is free. I had to buy an outboard Audio card for my laptop to make it work, but it was probably the cheapest way to do some more advanced testing than a simple DMM can provide.

You can get that software here.

 

[GeekGirl]

Is bass boost common? I understand this is a topic that's left up to personal preference, but is it used in practice? It's simple enough to put in.

What's a good cut-off frequency and level of boost

? I'm using Audio-Technica ATH-A900's (closed circumaural). The PPAv2 schematic parts give a 6 dB boost with Fc = 220 Hz (tangent's on-line calculator, Micro-Cap is slightly different).

 

[Clutz]

I'd say it's reasonably common. For the $20 for the POT and cost of the capacitor, I'd say it's worth it. I'm using a 0.022uF cap instead of the default value suggested because I originally found the original value didn't provide enough of a boost for my liking at the time.

Man, I want to build a STEPS and an M^3. I'll lay off the M^3 until after my source improves substantially..

That's where I should be putting my money.

 

[NeilR]

I use Senn HD-650's at home and Etymotic ER4-S on the road. I rarely use bass boost with the HD-650's; I always use it for the Etys. I'm trying to say that it is very dependent on your headphones and what you like. I think most people, for example, find the Etys to be a little thin on the low end.

You might want to socket those two caps and buy a couple different values to experiment. That's what I did for my Pimeta; then when I built an M3 I knew what I wanted because the results are about the same in any of these feedback loop bass boosts. I used .22uf caps. That gave me a very low shelf, which is what I wanted because I was happy with the Ety's until the very low registers. That is a very subtle bass boost and some people that used a similar value have commented that they weren't even sure at first that it was working. Meaning it shouldn't make the bass go BOOM BOOM like consumer gear bass boosts sometimes do. You really need to try this yourself, though, to see what you like.

 

[GeekGirl]

Thanks, I'll experiment. It's a tough call to fine-tune, so I'll just get it roughed in. Too light a capacitance moves the corner frequency up into mid-range. Too much resistance increases both gain (shelf) and group delay to a few mS on the low end.

Time to pick values (0.1 uF / ~50K pot) and move to the active components. Was fun using Micro-Cap for the analysis, though.

 

[kklee]

I just finished building a PPA V2 using the recommended parts on Tangent's site. There were two areas where I used the listed alternate parts. BD139/140 in place of MJE243/253 (couldn't find any at the time) and OPA627 (sound preference) in place of AD8610. The buffer bias is set to 20mA and OpAmp bias is 1mA.

The amp works great with higher impedance headphones (K701), but I would get a horrendous oscillation as soon as I used lower impedance (K81DJ, ER6i). Upping the buffer bias current to 30mA got rid of the oscillation, except for turning the amp on or off, where I would get a huge transient (offset/oscillation pulse > 2V!). I tried swapping out the OPA627 in the ground channel with an AD8610 and the amp settled right down (buffer bias @ 20mA).

I've read that the MJE243/253 pairs are recommended for low impedance, so I've got some on order.

I was wondering if somebody could shed some light on why I'm seeing this. Once I swap the buffer transistors out, will I be able to put the OPA627 back in the ground channel? I suppose I could just stay with high impedance cans for this amp, but I'm curious about this behaviour.

Thanks,
Ken

 

[Clutz]

I've experienced the same problem with my PPA and my AKG K24 or K26P headphones. I'm using the MJE243/253 combination, and I still had problems at the 20mA of buffer bias. in fact, I believe I also had difficulty at 30mA of bias.

 

[steinchen]

what output resistors did you use (R24/R34) ? If you picked 2R2 resistors it's a good idea to try 3R3 or 4R7 instead. What compensation cap did you pick for C6G (capacitance and type) ?

 

[kklee]

Quote:

Originally Posted by steinchen what output resistors did you use (R24/R34) ? If you picked 2R2 resistors it's a good idea to try 3R3 or 4R7 instead. What compensation cap did you pick for C6G (capacitance and type) ?

I used 2R2, I've got some 4R7 on order to try out. As for C6, I used the spec'd part (Digikey 338-1061). Would it be worthwhile to up it to 100pF?

 

[steinchen]

not to 100pF but up to the recommended 33pF, pick a silvered mica or C0G ceramic multilayer cap

 

[ppl]

Most instability issues with all versions of PPA's are the quality of the Input RCA jacks the PPA absolutely requires a massive chassis as a ground due to the lack of a Ground plane. 100pF is ok in the Ground channel with OPA627 in. The opa627 really don’t like to be operated at unity gain and has allot less phase margin than the AD8610. While using higher value Emitter resistors on the output gains stability it is not the best method because it increces the output impedance and Distortion. I have used 1.1 ohm emitter resistors with OPA627/OPA637 combo biased at 25 mA with complete stability on about 12 different Headphones.

 

[kklee]

Quote:

Originally Posted by ppl Most instability issues with all versions of PPA's are the quality of the Input RCA jacks the PPA absolutely requires a massive chassis as a ground due to the lack of a Ground plane. 100pF is ok in the Ground channel with OPA627 in. The opa627 really don’t like to be operated at unity gain and has allot less phase margin than the AD8610. While using higher value Emitter resistors on the output gains stability it is not the best method because it increces the output impedance and Distortion. I have used 1.1 ohm emitter resistors with OPA627/OPA637 combo biased at 25 mA with complete stability on about 12 different Headphones.

It's in the spec'd Hammond chassis, which is grounded through the volume pot. I didn't have any RCA jacks on hand (they're on order), so the input is currently using an isolated 3.5mm jack with a 5G iPod as the source.

I had an extra 10pF capacitor, so I tried tacking it in parallel with the existing C6 and found that the oscillations settled down a bit, but it still had an offset of 1.7V! I'll have to get a 100pF and try it.

The OPA627 datasheet states that its supposed to be stable at unity gain, so I'm still wondering why it only has problems with low impedance headphones. Shouldn't the buffer circuit be isolating the impedance load from the opamp?

...Ken

 

[ppl]

Quote:

Originally Posted by kklee It's in the spec'd Hammond chassis, which is grounded through the volume pot. I didn't have any RCA jacks on hand (they're on order), so the input is currently using an isolated 3.5mm jack with a 5G iPod as the source. I had an extra 10pF capacitor, so I tried tacking it in parallel with the existing C6 and found that the oscillations settled down a bit, but it still had an offset of 1.7V! I'll have to get a 100pF and try it. The OPA627 datasheet states that its supposed to be stable at unity gain, so I'm still wondering why it only has problems with low impedance headphones. Shouldn't the buffer circuit be isolating the impedance load from the opamp? ...Ken

No the first thing you do is hook up the RCA Jacks properly A SECURE CONECTION FROM THE INPUT GROUND TO THE CHASSIS IS MANDITORY NOT JUST RECOMENDED. Your grounding only at the pot is not what you were intructed to do on tangents site so why not wait intill you get your RCA jacks that are on order and after they are hooked up see if your instability goses away. The PPA is designed to work properly when built with the default parts and if it dose not first suspect builders error, Bad solder joints and flux left upon the PC board.. and again the Input ground on the PPA is not nice or desirable to be solidly grounded to the case. so ya you could use large caps in the ground channel and higher than recommended emitter resistors in the output to gain more stability than designed for however the PPA works quite well with your chosen Op amps if properly constructed and this includes removing all the flux from the PC Board. Remember the primary reason a DIY project or Electronic Kit don't work as designed is improper assembly

 

[NeilR]

Quote:

Originally Posted by ppl ...A SECURE CONECTION FROM THE INPUT GROUND TO THE CHASSIS IS MANDITORY NOT JUST RECOMENDED. Your grounding only at the pot is not what you were intructed to do on tangents site...

Hi ppl,

Your statement above got me curious because I cannot find any reference in the build instructions to specifically ground the input to the case. For my recent PPA build, I don't think I isolated the jacks because they are inherently isolated and I didn't have any problems that would have lead me to ground them. Just curious...

Regards,
Neil