Problem with the high hum in a tube amplifier for a guitar

[Thomasdj]

tube amp negative feedback

Hi Frank!

I wrote it wrong, the FET is connected at the mute buss at the "CH MSTR WIPERS" FET drain. I may have got this part of the schematic wrong. I've used six FETs (three FETs for the mute circuit in switching part schematic, and three FETs for the preamp pt2 and pt3) which could be only three FETs drawed two times in the schematic.

However, I measured the voltage at the gate of the FET in present and it shows 0v! Here is clearly a problem and I'll check if I've got the pin connections wrong. I've tried a lot of FETs with different brand so this could properly be the reason.

But, what do you think - is it three FETs drawed two times or six FET's? and would it matter?

Thanks

Thomas

 

[FvM]

tube power amp humming

Hello Thomas,

yes, that's what I meant with schematics misunderstanding. The FET's have been obviously drawn twice. Simply replace the FET at the switching part 1 with a wire.

Regards,
Frank

 

[Thomasdj]

shorting causing hum

I'll do that Frank!

Thanks

Thomas

Added after 4 hours 39 minutes:

Hi Frank!

The three channels are working perfectly now!!!

Now I'll focus on the FX-loop and then I guess the amp is ready for new tubes!

Thanks

Thomas

Added after 9 minutes:

Btw...what are the three FETs good for if they are on all time?

Is it a kind of slow "start/stop" safety process?

 

[FvM]

tube amp heater hum

Hello Thomas,

the FET's are muting the ouput (shorting the signal) for a short time when changing channels to suppress switching transients (clicks and worse).

Regards,
Frank

 

[Thomasdj]

bs170 preamp

ok Frank,

It was also my thought at the biginning. I measured the gate voltage (with my low-cost multimeter) when switching channels, but it didn't change! Now I've measured with the o'scope and see the changing. It drops the voltage to 0v which should open the FET and the signal should go to ground, right?

well, I can hear some noise (a tiny pop) when switching, could this be that the FET I use simply don't react fast enough?

If so, where to see in the datasheet, how fast the FET reacts?

Thanks

Thomas

 

[FvM]

amp hum not the caps

Hello Thomas,

the FET itself reacts much faster than you can ever perceive (< µs). The control signal at the gate is probably much slower. Some residual switching noise is tolerated by design, I assume. Also the higher on resistance of 2N5462 compared to J175 may play a role, but not necessarily.

For completely silent switch-over, an overlapped (with delayed channel switch action) soft muting (fade-out/fade-in sequence) would be needed. I don't think, that this effort would be appropriate for a guitar amp.

Regards,
Frank

 

[Thomasdj]

do tube audio amplifiers hum?

Ok Frank, I see..

I will finish this built now, put in a cabinet and try it out through the large speaker caps..

If I can't live with the switching noise I guess I'll try to install vactrol/LDR instead of RY2b (relay that switches between ch1 and ch2/3). This should silence the channel switching.

Thanks for your help..

Thomas

 

[FvM]

tube amp problems

Hello Thomas,

I don't know, how severe the switching noise is. You could check, if a improved mute control could reduce it. You can manually pull the mute bus to ground before switching the channel and release it later. If the delay causes an improvement, we may think in detail about a circuit modification. It could also be that the mute action itself causes a click. This effect should be treated different.

Regards,
Frank

 

[Thomasdj]

topology tubes hum

Hi Frank,

I just tried to ground the mute buss as you suggeted..didn't improve the noise. The noise is loudest (click/pop) when switching to ch3 from ch2 or ch1. Guess this means that Ry4 is the problem.

When switching from ch1 to ch2 I also here a click/pop but only the first time. Switching to ch1 don't click/pop.

I don't know how to improve this issue, other than using vactrols/LDRs. However this is a complete redesign of the switching circuit.

Another thing, pulling the gate to ground didn't mute the output completely. The signal just became remarkable lower.

Thanks

Thomas

Added after 55 minutes:

I have another question that need to be cleared Frank..

I now run the amp with four EL34 power tubes but I got a new set of power tubes waiting for the amp. It is a integrated quad of 2xKT88 and 2xKT66. I've modded the schematic so the amp got four individual bias adjustment pots (one for each power tube). However, I'm not sure how to place the new tubes.

The output transformer has two sides A1 and A2 on the primary. Should each side have 1xKT88 and 1xKT66 or should one side have only KT88's and the other have KT66's?

Do I have to be aware of anything here?

Thanks

Thomas

 

[FvM]

guitar tube amp design hum

Hello Thomas,

as I already mentioned, using original J175 would increase the attenuation in muted state. But I don't know, if the clicks can also go through the FB' path, that isn't affected by the mute circuit. Completely shorting the muting transistor DS can show, if a lower ohmic transistor would help.

The tube placement should be symmetric to my opinion, with a pair of KT66+KT88 at each side. The bias adjust may be a bit complicated, cause no individual anode current measurement is possible in the circuit.

Regards,
Frank

 

[Thomasdj]

amp hum problem

Hi Frank,


I've modified the bias supply so that each power tube has its own bias pot(see attached file). I've inserted four 1ohm cathode resistors to measure the cathode voltage. This works perfect with the EL34 tubes, but haven't tried with the new tubes yet. I guess the bias adjustment will still be tricky since the KT88 and KT66 draws different plate current.

It is not easy to short the DS since the FET is on the PCB. But I've tried to pull the 100 ohm resistor at the FET to ground. This did indeed decreased the click/pop volume.

The FX-Loop is bypassed by Ry10 so don't affect the problem now, but could when it's on. It also has its own mute FET.

Could I use a BS250 mosfet transistor with Rds on = 9-14 ohm ?

**broken link removed**

Thanks

Thomas

 

[FvM]

pinconnection wecp 20

The power stage circuit is clear so far. With help of the 1 ohm resistors, you should be able to adjust the bias currents. The guys who invented the KT66/88 combination may have an idea about an appropriate bias point. There are many possible settings, I think. Also the two types have different plate dissipation rating respectively maximum current.

If shorting the FET doesn't remove the switch click, a different interference coupling path may exist. (I assume, that you applied the short to the same ground that's used by the FET!). Then the FET on-resistance probably doesn't matter.

BS250 can't be used, cause it's an enhancement transistor while JFETs are generally depletion transistors, "on" without a gate voltage, "off" with a positive voltage. Also MOSFET are unsuitable, cause they have a substrate diode shorting positive Uds. Original J175 should be preferred (if the the On-resistance is an issue, which hasn't shown clearly yet, to my opinion).

Regards,
Frank

 

[Thomasdj]

ground star tube amplifier

Hi Frank,

I will set all tubes to 70% of plate dissipation. Another thing will also "move" the bias current, - the rectifier select, since the plate voltage will be lower when using rectifier tubes. I use the GZ34 instead of 5U4G, since the voltage drop is only 10-20v with the GZ34.

If I short the signal to ground (right at the FET) then the click/pop almost disappear. So, the ON-resistance does matter, but something else comes into the signal. I'll see if I can find some J175.

Here is something wild guessing!! What if I connected another FET in parallel??? - would this work the same way as resistors in parallel?

It could also be that the realys are switching slower than the FET, am I right?

Thanks

Thomas

 

[FvM]

guitar buffer bf245 bias

Hello Thomas,

The relays switching slower than the FET allows the FET to act before generation of switching clicks/pops. The mute bus is then released delayed by a capacitor.

J175 in TO92 case is on stock at least at farnell and digikey.

Regards,
Frank

 

[Thomasdj]

increasing hum in my tube amp

Hi Frank,

Yes, I'll order the JFETs and see if this minimize the problem..

Thank you for your help!

Thomas

 

[Thomasdj]

vacuum tube hum when cold

Another question Frank,

Would this FX-loop mod work? to switch between serial and parallel effect loop..

 

[Thomasdj]

tube amp connecting heaters ac dc

Here is the link to the panel meter:

https://www.circuitspecialists.com/images/CX101A&B&BG.pdf

I only need the 0 - 200mV DC range so I wouldn't need any additional resistors, right? would that mean that I will have to connect pin 2,5 and 6 together as ground signal? and 7 and together as V-in source??

Thanks

Thomas

 

[FvM]

tube amp hum input

Hello Thomas,

the DVM module has an internal connection of COM to mid supply (with a voltage divider respectively a reference voltage). I guess it is using an ICL7109. If the module manufacturer doesn't give more information, you can understand the supply voltage requirements from a 7107 datasheet. **broken link removed**

Cause the DVM chip has a differential input, a differential voltage divider allows a grounded supply, as shown in the left schematic. But it doesn't work for 200 mV input.

Alternatively, you can use a bipolar supply and a differential input without connecting COM. The analog input common range is defined such: (V- + 1V) < VlN < (V+ - 0.5V). This means a -3/+4..+6 split supply should work. But it's about 20 years since I have used ICL7107, I don't remember every detail.

P.S.: I don't see clearly the purpose of the above modification. I understand, hat the mix control allows to vary the proportions of effect return and original signal. The additional switch is for selecting 100% original/ 0% effect return? But the footswitch controlled FET circuit has excactly the same purpose.

Regards,

Frank

 

[Thomasdj]

tube amp hum with cable only

I'll try the split supply

Frank, I think the mix pot is to mix the original signal (dry) with the effect return signal (wet)..if cutting the original path, wouldn't I get the wet added signal only?

Thanks

Thomas

 

[FvM]

bf245 pre amp

if cutting the original path, wouldn't I get the wet added signal only?

O.K., that's a plausible purpose.