or Connect
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › A Very Compact Hybrid Amp
New Posts  All Forums:Forum Nav:

A Very Compact Hybrid Amp - Page 63

post #931 of 2213
The idea of tolerance/heat issues in the OB is interesting, since we're in a complex system here - wonder how that relates to slightly different tolerances in the splitter circuit.

Temporary heatsinks (alligator clips?) on to-92s in the OB? If they got replaced, it seems maybe that it's something upstream , perf CFCubed's experience replacing lots of things (although odd as the splitter feeds both buffers).

It's my understanding that Runeight has been swamped lately, I'm sure he'll weigh in soon.
post #932 of 2213
I was trying to check out the tube servo. I am not sure if I am understanding this, but on pin 6, I was getting @78vdc whith spikes up to 108v, triggering the e12. This must be it, right? This would point to a input issue, not the buffer. D1R? I did earlier swap out ICR. Something else I am missing?
post #933 of 2213
Thread Starter 
Gentlemen, my sincerest apologies for being awol. There are many nice things about a consulting practice, but when the chips are down the customer gets the first call . . .

Reading through I think that the only unsolved problem is tacitapproval? Is this right?

If so, tacit, that pin 6 voltage spike would definitely trigger the e12 since the spike will pass right to the buffer. It's hard to say exactly what the problem is, but it could either be a bad D1R or it could be an intermittent connection.

You might try very carefully inspecting around the tube servo for the right channel. Even flexing the board a bit to see what happens. If you can't trigger it this way, check out the opamp socket and opamp fit. If the opamp flakes out or loses even one active pin the spike will happen.

If you don't get anywhere with these steps then replace D1R. But my sense is that if it were bad it would just be bad. Unless it is cracked or something. But electrically it would either work or not.

Hopefully getting the tube stable will fix your output problems.

Let us know.

Is there anythine else pending? Thanks to everyone for holding down the fort while I've been gone.
Dr. Cavalli gained notoriety with his first DIY amplifier projects. His success has blossomed into Cavalli Audio, a world leader in amplifier design.
Reply
post #934 of 2213
runeight -

I was wondering if we can adjust the led current from the e12 to light the tube better. Something like 10-15mA
post #935 of 2213
Thread Starter 

Temperature Issues

There have been several comments about BJT temps and its effects.

BJTs are quite sensitive to temperature and have a positive coefficient, meaning that the hotter they get the more they conduct. Mosfets have a negative coefficient meaning that the hotter they get the less they conduct. This is why many designers prefer mosfets for O/P devices because they won't thermally run away.

Any BJT complementary buffer is subject to temp issues. If one side of the pair is heated or cooled the offset will stray to one side or the other. This effect will be pretty fast.

You don't notice this on many amps because there is no offset protection circuit. The DC offset slops around but nothing happens to tell you this. So you go merrily along thinking that there is nothing going on but audio.

Having an e12 tends to make it look like there is a problem here when, in fact, the amp is doing what most other amps with similar O/P stages are doing.

However, what happens to the DC offset also depends on the exact circuit used. For example, in a design like the PPA where the gain stage is an opamp and the discrete buffer is in the opamp feedback loop, the opamp will zero the DC offset quickly and there will be no drift.

The extra challenge here is the hybrid design where the buffer must self zero and, therefore, must have a longer zeroing time constant so that the buffer opamp does not respond to low freq audio.

In this case the buffer is subject to being rattled by input spikes and fast changes in the O/P devices. It will eventually pull back to low offset. The e12 is set to accomodate slightly large swings before triggering to handle this design. And, I'm pretty sure that when the design is working right (correct components, nothing intermittent, no spikes coming in from the source) you won't have e12 trips.

I have designed an amp where the tube is not capacitor coupled and is in the DC feedback loop of the amp. It's the EHHA for those of you who would like to try an embedded hybrid that is totally DC coupled.
Dr. Cavalli gained notoriety with his first DIY amplifier projects. His success has blossomed into Cavalli Audio, a world leader in amplifier design.
Reply
post #936 of 2213
Thread Starter 
Quote:
Originally Posted by rds View Post
runeight -

I was wondering if we can adjust the led current from the e12 to light the tube better. Something like 10-15mA
rds, the LED resistors are 2k and are essentially strapped across 20V. So they should be passing about 10mA now. But you can definitely burn more current. Try 680R to get about 15mA. 510R to get 20mA. I don't think the LV supply will notice this change and the e12 control transistors should be ok.
Dr. Cavalli gained notoriety with his first DIY amplifier projects. His success has blossomed into Cavalli Audio, a world leader in amplifier design.
Reply
post #937 of 2213
Thread Starter 
Correction. Bad math.

Try 1k3 for 15mA and 1k for 20mA. Sheesh.
Dr. Cavalli gained notoriety with his first DIY amplifier projects. His success has blossomed into Cavalli Audio, a world leader in amplifier design.
Reply
post #938 of 2213
cool, thanks runeight - I'll give that a try.
post #939 of 2213
@tacitapproval.

108v? Like runeight said, either it works or it doesn't. It could be an intermittent connection, but also I'm wondering if it's a PS thing. If you can, reflow that section. But after that, and examining for cracks and flexing the board, check your B+. Does it spike at the same time?
post #940 of 2213
Ok, I reflowed all the joints on the right side of the input section. I am now getting TB+ 108V, pin 1 78v, pin 6 34v? Shouldn't this be 78v as well? The good news is it seems to be stable and music is playing through both channels. That pin 6 reading is confusing me.

edit -- I swapped to another tube and now getting 78v. But why would that other tube (12bh7) sound fine at 34v on pin 6, let alone why was it reading that?

edit -- Now it is at 46v and e12 trips are back. Something is obviously still wrong. As usual, I'm confused.

Holland - B+ seems to be 108v consistently. It moves around from 107.5-108.2, or so, but doesn't seem to spike when the e12 trips.

Also, I measured D1R against D2L, and they seem to be about the same.
post #941 of 2213
yep.it should be reading 78Volts..looks like you might have a problem with your Q3P...
post #942 of 2213
Quote:
Originally Posted by runeight View Post
I have designed an amp where the tube is not capacitor coupled and is in the DC feedback loop of the amp. It's the EHHA for those of you who would like to try an embedded hybrid that is totally DC coupled.
God is it a beautiful design...absolutely love it..just the heater circuit left to finish up on mine
post #943 of 2213
Quote:
Originally Posted by tacitapproval View Post
Also, I measured D1R against D2L, and they seem to be about the same.
I don't understand this. Do you mean D1L? What do you mean by measured against?

You measured B+ via TB+? What about on the other side of D1R? What are the voltage drops across D1L and D1R with your various tubes?

This tube you are having problems with? Is it the same one that has always given you problems? A 12BH7 is it? A different 12BH7 works? Am I understanding your problem correctly?
post #944 of 2213
Whoops, yes I mean D1L. I meant that I was getting similar readings between them on the diode setting of my dmm. Under power, I am getting 78v out of D1L and D1R. It is not one tube that I am having problems with, I have tried several different varieties.

Ok, now I am getting a steady 78v out of pin 6. This is confounding. Hopefully, it was just a bad joint and my various reflows got it. Of course, I have thought it was stable before.
post #945 of 2213
reflow it again, for good luck. Look at the joint from the top. There should be some solder coming through the joint and holding.

As sachu said, there is a transistor between TB+ and D1R. That is Q3P. When you measure the low voltage on pin 6, measure the other side of D1R and TB+. It could be a cold joint on Q3P as well.
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › A Very Compact Hybrid Amp