Doug's AB763 project amp - bias surge question

[NeilG]

Hello to all - my first post.


Had been revisiting a previously completed 2 channel project to address one final issue. The amp is fully functional however when setting the bias current, I noticed that the bias current surges on both power tubes once the amp is warmed up, when switching from standby to full on. Both tubes draw over 80 ma's before settling to my target of 36 ma. Doug's notes suggest this is not desirable.


I did try increasing the value of the resistor from 470 to 670 ohms, but saw no improvement.


I've checked the wiring and components multiple times. The voltage at pin 5 of the output tubes is apprx -46 relative to chassis ground. Will dip to apprx -28 on switching to on before returning to -46.


Any suggestions are appreciated.


Happy Holidays and thanks in advance.


All the best ...... Neil.

[HotBluePlates]

... The amp is fully functional however when setting the bias current, I noticed that the bias current surges on both power tubes once the amp is warmed up, when switching from standby to full on. Both tubes draw over 80 ma's before settling to my target of 36 ma. ... The voltage at pin 5 of the output tubes is apprx -46 relative to chassis ground. Will dip to apprx -28 on switching to on before returning to -46.

Suggestion: Don't measure bias at the instant you flip from Standby to On. 

Okay, that was a little flippant... You have probably 2 things happening at once when you flip to On: All tubes start at some higher-than-normal B+ and plate voltage (because none of the tubes are drawing current and sagging the supply before the amp is in the On position), and you have a brief, large positive pulse from the Tremolo circuit as a secondary result of high supply voltage phenomenon.

I think this is probably unavoidable in this circuit. Or at least, difficult to design-away. I also seriously doubt you're doing the output tubes any harm because they only have the low pin 5 voltage/high current for what, a part-second? The tubes will tolerate this well. I never even thought to check amps for this issue when I was installing Hoffman AB763 boards for customers, and never got any calls/emails about damaged output tubes.

I'd also recommend against obsessing over bias (not that you are yet). If you kept the meter attached for a long time (hours, days), you'd see the tube idle current likely drift a lot in that time. If the tubes are new/unused, probably more often and wider amounts than if the tubes have been used for a period of a few hundred hours. This is also a known, normal phenomenon, and the principle reason a burn-in was advised before making critical measurements on tubes (the other reason was that tubes will either suffer infant mortality during burn-in, or last a very, very long time, so the burn-in period weeded out the tubes most likely having defects).

There are ways (circuits) to stop the idle current variation, or to clamp the idle current to some maximum value. The simplest way is to use cathode bias instead of fixed bias (more current creates more bias voltage, which reduces current). For fixed bias, servo and/or timing circuits can be used, but adding them seems anathema to a simple hand-built amp.

I'd just play it if the amp is sounding good and otherwise functioning properly.

[sluckey]

The voltage at pin 5 of the output tubes is apprx -46 relative to chassis ground. Will dip to apprx -28 on switching to on before returning to -46.

That's what is causing the bias current surge but I've never seen that before. Pull the output tubes and monitor pin 5 while switching from standby mode to operate mode. Does the voltage still dip?

[EL34]

The bias voltage should stay very steady when flipping the standby switch.


Show me a schematic of how you are getting your bias voltage.
Where is your source AC?

[NeilG]

Thanks to everyone for their replies.

I did compose a detailed reply this morning however, it seems to have gotten deleted, so I will re-create it.

The project amp was wire exactly per Doug's design with any mods highlighted.

For the bias circuit, I increased the value of the 10 KL pot to 15 KL, the bias resistor from 470 to 670 ohms - this was done after reading the notes regarding limiting the current surge, however, it did not work.

I added Sluckey's AB763 Tremolo Mods, 1 thru 5, including the LED mod. An good improvement.

I also added an IN4007 from lug 1 to 3 on the Intensity pot per a Princeton Tremolo mod I found - can't remember where I got this mod from but it eliminated unwanted "pulsing" in the tremolo.

The bias voltage is drawn from the red/blue stripe  Power Transformer lead tap, per the schematic.

The Power Transformer is a Heyboer HTS 9409 - I am a little suspicious of this PT - wondering if it's undersized?? I also have a new Hoffman 022798, but could not mount it as the chassis slot was too small.

I measured the voltages at pin 5 with and without the Power Tubes, with the same results - standby to on and back - 46 to 53, settling to 46. The dip to the 20's, settling to 46.

Swapped the P-Tubes with the same results. Looking at the meters on my Alessandro bias unit, the current for V8 surges past 100, whereas V7 only surges to about 54 ma.

[sluckey]

I measured the voltages at pin 5 with and without the Power Tubes, with the same results - standby to on and back - 46 to 53, settling to 46. The dip to the 20's, settling to 46.

I highly suspect a wiring error. Can you post some hi rez pics?

[NeilG]

Here are some pics -

Please tell me specifically what I should capture and submit.

Thanks ..... Neil.

[NeilG]

Another pic

[EL34]

Neil,
The wiring is everywhere and like spaghetti
No way I can tell anything

I can't say this enough times

When you get done following all wires and checking all component values, everything should be highlighted

[EL34]

I can easily follow the wires in this image

I would start from scratch and re-wire the whole amp like this

[NeilG]

Thanks Doug, as you figured, I couldn't make a living at this and I'm okay with that. No intention to waste your time, so sorry for that. For reasons I won't go into on the forum, these tasks are a challenge for me. The amp still sounds really good, so thanks for this.

Merry Christmas ....

[EL34]

I would just do a little bit or re-wiring at a time until you get the amp re wired
I am sure you will find the problem once you have finished


Otherwise, the problem will still be there


Shorten everything as you go
There's no need to have that much wire

[sluckey]

I can't stop looking at the pic of your bias diode. I know that from everything you've said it should be connected properly. But there is a glare on the diode that makes it appear to be reverse connected. The banded end should connect to the 680Ω resistor. Can you verify this is true so I can quit staring at that pic?

[NeilG]

Sluckey, finally able to restart the project, I've decided to start from scratch. Ordering the board with turrets today.

I did replace the diode and it's properly oriented, so this was not likely to be the problem.

Version 2 now in motion.

Again, thanks to everyone for weighing in.

Neil.