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

[Thomasdj]

ground loop amp tube

You're right Frank!

I couldn't see the light from the LED simply because of the daylight or room light.

I measured from -3v to pin 18 (channel 2) at 564 and the meter shows +3.25v in "on" position and 135mv in "off" position. Since the voltage is ok, could this be some kind of lack of current?

Best regards

Thomas

 

[XNOX_Rambo]

my tube guitar amp has ahum in it when cold

With that 3.25V and a voltage drop of 1.7V over a (red) LED and 470 ohms you should get about 3mA through the LED. Which is fine.

If you on the other hand have a blue or white LED you won't see much... :wink:

/Ram

 

[Thomasdj]

guitar buffer bf245

Hi Ram,

I simply don't get it..the LED's I use are actually blue..I've also tried red LED's (3mm and 5mm) but it doesn't change anything.

Here is a link to the Blue LED's datasheet..mark that they are not different from the red LEDs

https://www.futurlec.com/LED/LED5B.shtml

Thanks

Thomas

 

[XNOX_Rambo]

fix 100hz hum

They must have messed up that web page - because if you look at the 3mm blue LED the values are more normal:
https://www.futurlec.com/LED/LED3B.shtml

A Vf of 2.0 for a blue LED would have been a world sensation... :wink:

Man, this project must be cursed!

/Ram

 

[Thomasdj]

eliminate hum from tube heaters

Hehe..yea it is indeed cursed! (because it's the first known clone on the net :twisted: )..however, I've tried all kinds of LED's..3mm Red, yellow and green, round and square..and nothing is changing..

I use the 74ALS564, don't know if that could be the problem. Tried with 74HCT564 but then the circuit does not opperate propperly at all..

I use Blue, white and purple LED's for the indicators at the frontpanel of the amp (switching matrix LED's) which works 100%.

Frank wrote that the 1k at pin 11/564 could be pulled up to +2v rather than pulled down to ground (-3v), would this have any influence on the three output of the 564? If not, where do I get more current!? :?

Thanks

Thomas

Damn I am feeling new why not bypass the 470 ohm??

Problem is fixed! everything seems to work now, tanks for your help!

Now I'm just waiting for the log pots to arrive..

Best Regards

Thomas

 

[XNOX_Rambo]

100hz hum amp

When pin 6 on 7420 goes high a pulse through the 0.1µF capacitor will trigger the clock input at pin 11 on 74564,
which is edge triggered on a rising flank.
The resistor will then discharge the pin to ground, so in my view it looks right.

The pulse at the clock input must be over 2V for the 74564 to latch the output pins 17, 18 and 19
at the states of the input pins 2, 3 and 4.

You could try and increase the capacitor to e.g. 1µF to give the pulse more oomph - but as the switching works
that doesn't seem to be the problem.

Could it be something simpler? Like the 470 ohm resistors being 4700 ohm? :wink:
Because if you have the voltage you will get the current.

/Ram

 

[Thomasdj]

should my tube amp have a hum

Exactly Ram!

I figured out, I haven't checked the res values but I think I'll bypass them..it all seems to work now

Added after 1 minutes:

No, Values are Ok

I think it's because of the "ALS" , but not sure

Thanks!

 

[XNOX_Rambo]

channel switching ldrs tube amplifiers

Ahh, now I see a problem...
The 74ALS564 states that the Ioh, High-level output current, is max 2.6mA.
On the other hand the sinking current can be as much as 24mA.

A HC type has a symmetrical output, i.e. it can sink and source an equal amount of current.

I found a comparison here: https://www.fairchildsemi.com/an/AN/AN-319.pdf

I would go for HC parts all over...
Edit: But it is probably enough to exchange the 564.

/Ram

 

[Thomasdj]

open circuit amp hum

Yep!

Interesting comparison table! I think I'll just replace the 470 ohm resistors with someone lower..or just run the cathodes directly to ground..

btw..I'am actually impressed how well the circuit responds/opperates..the switching is fast and works great. hopefully this don't change when I connect a 6 meter cable between the power supply, signal and the footswitch (guess a capacitor would improve stability if this become a problem).

Thanks

Thomas

Added after 14 minutes:

...it already got a cap :? ,,a bigger one then :wink:

 

[XNOX_Rambo]

dc twisted tube amp

A wild'n'crazy idea I've got - well in line with the footswitch design :wink: - is to use unused outputs to double the 564 sourcing current.

I.e. connect pins 5 and 2 - and pins 16 and 19. Do the same with the other two: 6+3/15+18 and 7+4/14+17.

/Ram

 

[FvM]

dc filement hum

Hello Thomas, hello Ram,

I trusted in your previous statement, that a HC564 was used and wondered about the low output voltage driving the LED. For TTL, LS or ALS it is regular, they have a rather low source current (positive) and higher sink current (negative). I didn't think about, but this already clarifies that HC type is used in the original circuit. Bipolar logic as ALS has also a unwanted current consumption.

A HC or HCT is also able to drive a blue LED, but it should use a current limiting resistor.

The other problem is with latching of switch pulses and mixing different logic families. If I understood the circuit right, a positive edge at 564 clock input is now generated when releasing the switch. But the input signal isn't present then any more. It may work with a combination of HCT20 and a HC564. A very strange design, to my opinion, if it's actually the original circuit.

Or, the 1k resistor is simply drawn incorrect. Then operation should be possible independent of logic family. Hopefully there is no false switching due to contact bounce on switch release.

To evaluate the circuit operation, it may be meaningful to generate the latching pulse manually. Also an oscilloscope (a digital storage type would be good) can show what's actually going on.

Generally, there are so many possibilties to find a functional circuit, I would decide normally depending on which parts are in stock, respectively used in other instruments. It seems to me, that a cross coupled 3x3 NAND circuit (with 74HC10, an extended version of the usual 7400 two input latching circuit) would also do in this place.

P.S.: I see, that a pure combinational logic latch circuit would be more complicated than a 3x3 NAND.

x <= (NOT a OR x) AND b AND c;
y <= (NOT b OR y) AND a AND c;
z <= (NOT c OR z) AND a AND b;

Regards,
Frank

 

[Thomasdj]

what causes a tube amp to hum

Hi Frank/Ram,

I've tried to use a HCT564, but this does not work with the 7420 and 7406 neither with the 1k res pulled to positive. However, I replaced the 470 ohm resistors with 47 ohm, and everything seems to work propperly now. The LED's light (not super powerfull, but fine as indicators) and the channels are switching perfectly.

This is not the original design, but I think it's very close. Haven't tried the FX-loop and the Solo-function yet..but I'll do that later and post the results..

Thanks

Thomas

 

[XNOX_Rambo]

amplifier gain causing hum

Hi boys,

Well, from this we learn that designing by trial and horror with digital chips in an analogue way will
get you into all sorts of trouble.
Especially if you are going to make more than one unit... :wink:

Cheers!
/Ram

 

[DrBob13]

tube amp humming problems

Hi All.Thomas have you thought about using a small bipolar transistor to switch the EQ LED, this will buffer the output and guarantee that the LED will always light eliminating the low voloage problem with blue LEDs. All you need is fit a small switching transistor, the same type as you have used for the relay switching will be ok, feed the output currently connected to the 470R resistor via a resistor between about 1 and 5k to the transistor base, and connect the LED/Resistor between the 5V rail and the collector and then ground the emitter. It may be necessary to fit a 470R to the 2V rail for the circuit to work as it currently does, I have not thought too deeply about this.
You could do something similar for the Ch1, CH2, CH3 LEDs as well to ensure that they are independant of the logic family used, 74HCXXX should be fine incidentally, like RAM says, you can store up a whole load of future trouble if you design for production by trial and horror with a marginal circuit.
Good luck
Best regards,
Bob.

 

[FvM]

ground loop tube amp

Hello Thomas et al,

a redesign for production purposes would clearly use a microcontroller as a true single chip solution (apart from a few resistors and capacitors). But's an old design with several non obvious flaws. Digital circuits are really not the strong point of Mesa Boogie guys...

Buffer transistors are a good solution. I still think, that 74HC564 (or HCT) should work somehow, but I can't debug the circuit from a distance. Fortunately you found a way to operate it so far.

Regards,
Frank

 

[Thomasdj]

tube amp high frequency oscillation

Hi All,

By now the only thing I've changed was the 470 ohm res to 47 ohm. I run purple, blue and white LEDs for Ch1/2/3 and it works perfectly!

Bob, I actually did thought about using transistors but since the 47 ohm solved the problem I don't think it's nessecary. But, I don't now yet how the circuit reacts when using a 6 meter cable between the amp and the footswitch, but I'll get back with the results.

Next issue is the Log pots which I expect arrives today/tomorrow! After replacing the pots I'll focus on the FX-Loop and the Solo-function which I haven't tested yet.

I'm not sure how this works but I think the Output pot works as an overall master volume when the fx-loop is on. Am I right here?

I'm also drawing the front/rear panel, very nice! and the cabinet for the amp. So still have some work to do before I can test it completely (the amp isn't shielded before it's built in the cabinet a.s.o.).

btw...I also made some nice temperature controlled fan circuits, to cool down the chassis and blow away the heat from the output tubes (It gets damn hot!!).

Best regards

Thomas

Added after 2 hours 38 minutes:

Hi Frank,

I was thinking about what you said in the begining, about the hum issue and DC heaters. It could be vary awesome if I could quiet the beast even more than it already is. It have some/too much mid frequency noise response which I can regulate with the tone controls. Have in mind that this is observed when the amp is running near max, - beacuse of the wrong linear master volume pot.

I've made a drawing showing the current heater setup in the top and my question (future state) at bottom. would this mean anything to have two bridges? because, I cannot get to the heater wires since they are lying underneath the large preamp board..

Thanks

Thomas

 

[DrBob13]

dc filament filtering tube amp

Hi Thomas,
Personally I would have connected the low voltage supplies directly to the transformer with their own dedicated pair of twisted wires, you seem to have a bit of a daisy chain there. Otherwise it looks OK.
Try that before you go to DC. The mid range noise is very unlikely to be coming from the power supply, that could be due to a noisy tube or noisy component, cold joints etc, some tubes are more noisy than others, noise cannot be eliminated completely from a tube amp, you will always get some shot noise in tubes that is caused by the electrons hitting the anode, the noise from the low level stages although usually less than from the power tubes does get amplified up by the later stages. I used a purpose desiged switch mode PSU to provide an all dc solution for my last amp, it works brilliantly and is very quiet but is a complex beast.
You are doing well, it could be much worse, these things always happen with a new build.
Cheers,
Bob.

 

[Thomasdj]

causes of hum in a tube amp

Hi Bob!

Thanks, I see what you mean! I did the "daisy chain" only to minimize the number/lengt of heater wires since the low voltage sypply PCB is mounted "on top" of the preamp board and close to V1 / far from the power transformer.

I'll try too connect it seperatly when I get the log pots..

I am aware of noisy tubes, and in fact the tubes I use could be noisy (they're used and old). I've got a new set of tubes on hold, but I'll wait to the amp is 100% stable!

I only want to see if DC on heaters makes a difference..

Best regards

Thomas

 

[FvM]

tube amp problem

Hello Thomas,

the daisy chain is O.K. as long as the voltage drop along the wire isn't two high.

I would test a DC filament supply with a laboratory supply first. You need 1.2 A for the four preamp tubes. Bob is right, that not all kind of noise can come from AC interference.

Regards,
Frank

 

[Thomasdj]

how to grounding problems a tube amp

Hi Frank!

Thanks, I haven't got any wallwart supply with that power rating. But I'll try to make the circuit on a test board and measure before connecting it to the tubes.. Iwas thinking of using the same design as the mesa "low voltage supply" circuit. just with one 4700uf filter cap..would this be OK?

Best regards

Thomas