Tube head crackle and pop after warm up

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Would you please measure and post the output tube plate voltage, screen voltage and bias voltage. I would like to know how hot Dean Markley is running those tubes for a claimed 120 watts.

Hi,

Yes I plan to note PSU voltages. Contrary to my previous remark that bias didn't appear to be adjustable, it is adjustable, however not individual tube bias.

Which brings me to thinking, what bias current I should be expecting?

Referring to datasheet (attached), I could be either running AB1 or AB2 PPP, and bias values are for a PP pair.

So approximately 35mA to 50mA per tube at a guess.

What will be more interesting to find out, is what Voltage the plate and screen is supplied.

Judging by the series pair of 100uF 350V capacitors for the bulk reservoir, I'd guess that B+ is more than 500V.

Soon as I know I'll post some results, though I'm mindful that caps need changing (probably) and I should also probably do those first (eliminates one potential source of trouble)
 

Attachments

  • 7581A.pdf
    843.3 KB · Views: 29
Rayma,

Yes Cathodes a grounded to valve socket, all commoned with wire links, then single ground wire to PSU PCB.

However, I have picked up something. I dont think it's relatedto the popping though. The two 100R resistors on the end of the heater chain are only 1/4W Carbon, not the 1/2 W unless specified (on the drawing posted earlier)

I can confirm that drawing part no. matches the PCB part No. 550-056
 
Are the 100R heater virtual center tap resistors good? I would suspect AC hum from either one being open or cracked, but not the sounds you described.

The only one of my amps with a virtual heater ground has 1/2W 100R, but I've seen carbon comp 1/4W used for these in a number of amps; I've been told that these are sometimes sized at 1/4W to act as "fuses". I think properly sized flameproof resistors might be a wiser choice instead, but someone smarter than me may chime in.

Have you changed plate or cathode resistors, or any other parts yet, or are you going to follow the troubleshooting suggestions first? Sounds like you've already cleaned tube pins and checked sockets (re-tensioned?), so it might be better to diagnose first.

Regarding bias, might be best to follow the manual if you have or can find one, or use an online bias calculator after measuring, not guessing, plate voltage (which will change with bias setting).
 
Hi,

No I haven't changed any parts yet. I was tempted to change out old or suspect parts. So far I haven't found any other than a known area that gets hot and browned board (zener supply) and I have parts on hand to replace and uprate.

I figured if I left as is, knowing it works, so I can try and track down the issue.

I had a good poke about in the case before I made live again, as it had been sat unplugged for a day or two. Besides some weary looking, mangled legs on these 100R resistors, nothing is jumping out at me.

I have to wonder how I measure quiescent current for bias adjustment though, seeing as all cathode are tied directly to socket then star pointed, and a single ground return. It's going to take modification (ok that's a given), but in the most cramped tight spot!

Who ever decided to wrap the anode wiring around the PSU PCB and out to the OPT should've been sacked. The heater wiring is sloppy too. These amps are renowned for being hand built, and you can tell.

Re tensioning sockets? Well I dont know what that is, or how it is done. I just cleaned off the preamp valve pins, which were slightly tanned, and darkened up by the tube base. If I had thought about it, I wouldve used a flux pen, as it looked a little like blued metal tarnish.

I had the amp on for 20 mins, and no pops, then realised perhaps I need a cable plugged in. Cable plugged in an still no pops.

With Rayma suggested load jig (8R load shunted by series R and speaker) Albeit with a somewhat lower R of 470R.
Kudos for that idea!

So I waggle all potatoes* on the front, pull the master volume up halfway through the crackles and waggle away. Seems I may have sorted the pot crackle temporarily.

* autocorrect did this but it's hilarious and deserves to be left in

Still no pops....

I have closed curtains to search for red plating but its bright here this week. I see some blue glow, which I dont think is a problem. I haven't measured voltages yet, I need to find my clip on leads.

For now I've left tinkering, to resume when it's dark and I can check the for red plate more easily.
 
Still looking good :cheers:

Not quite visible in the darkened photograph is slight red plating on all 4 valves.

So I think a re cap is in order.
The 18V zeners will get changed for 1N5356
I have MIEC 100uf 450V electrolytic and a Chinese one, so I may have to source a decent pair.
I have the 33uf 450V in nice Cornell dublier, and I think I have the 22uF in 500V MIEC also.

Then maybe try to bias a bit cooler.

All except the 33uF are 105*C and all are updated un voltage spec. The originals (Illinois Capacitor) are, I think a still available, though it took me a while to find them and they're expensive shipping from the US!
 

Attachments

  • 20190705_204810-1382x1843.jpg
    20190705_204810-1382x1843.jpg
    479.3 KB · Views: 93
  • 20190705_204348-1843x1382.jpg
    20190705_204348-1843x1382.jpg
    142.6 KB · Views: 93

PRR

Member
Joined 2003
Paid Member
...I could be either running AB1 or AB2....

If there is not a *fancy* driver (rare), then it can't be AB2.

...wonder how I measure quiescent current for bias adjustment though, seeing as all cathode are tied directly to socket then star pointed, and a single ground return.

There's Bias-Rite and similar, socket adapters which break the cathode path for metering. But IMHO it is worth putting 1-Ohm resistors in permanently.

Measure plate-cathode voltage, figure 70% of rated Pdiss, and bias near there. There's no magic math here because there really is no optimum. But this is safe and works well for all conventional fix-bias guitar amplifiers.
 
PRR,

I thought I could break the cathode connection to socket and install resistors there (10R but perhaps 1R is better)
Thanks for the clarification, showing my lack of experience with valves, class AB1 it is! :cheers:

The only missing info is OPT Z (I assumed I would measure plate volts and draw the load line for some % Pdiss, but I guess incan just do the math and skip the load line.)

The datasheet shows Vp 450V, Vg2 400V, zero signal current 116mA for a pair. Without measuring the voltages, I reckon this is close to where the tubes lie, looking at the component ratings.

So at 70% by loose reckoning, that's 0.7*(116/2) bang on 40mA.
Which is one of the numbers written on the tube! I assume they were matched at 40mA

Wiseoldtech,

Red plate is bad and I understand it can reduce output, valves can be damaged. I've not a huge amount of experience with valves (besides this amp and a small amp I'm building, nothing) but I would say the red plate was very very slight in a darkened room.
I think/hope, I should be ok when I re bias.
 
Last edited:
How hot is HOT?

I went around the PSU board today taking voltages.

I think what I am seeing is the result if a somewhat high supply voltage 245V, and perhaps a transformer modification, or replacement at some point in the amplifiers life (I assume a 230V primary to replace a 110 or 115V)

Mains input 245V 50Hz
Heater 6.4V
HT WDG 360V
(Slightly high)

Here's where things get kooky

Referring to the CD120 schematic:

J = 482Vdc, mid point of 100uF caps 232V
F = 480V
G = 445V
Preamp output tap = 360V

E= -63V

So this means the plates are at about 475V, screens at 480V, phase splitter driver supply of 445V, 360V for V1 and V2.

These seem a tad high....

Assuming 230V transformer then (230/245)*482 = 450V.

A little better. Heaters would now be about 6V. Still in recommended operation range I think.

Is it usual to have screens at voltage so close to anode?
 
Last edited:
DJGibson,

Thanks for that titbit of information.

It would seem that it's not at all unusual for guitar amps to run the pentodes with screen volts similar to plate volts rather than say 20% less.

I wonder then, what difference to running in triode wired?

The only difference is g2 and anode aren't directly connected (isolated by supply choke)

I need to re measure the screen volts at the tube (rather than the PSU tap) just to be sure. Theres additional resistance in the screen circuit, so chances are that drops the supply voltage slightly.
 
Last edited:
When using triode mode you can get away with a B+ a bit above the usual (pentode mode) G2 screen voltage rating.

In pentode - as the tube conducts heavily on signal peaks, the anode voltage is pulled down and becomes much less than the screen voltage, as a result screen current peaks (and hence screen dissipation [volts x amps] peaks too).

In triode mode screen voltage and anode voltage are the same so the usual ratio between anode current and screen current stays the same. There is no extra peaking in screen current and so the screen can tolerate a bit more voltage without overdissipating.

Similarly for Ultralinear Mode.

Whenever an amp has had a red plating of output tubes experience, (run away anode current) it is a fair bet that it has had screen current peaks too, and those 470 Ohm 2W screen resistors suffer. They can be a source of crackles and pops. Check them with a multimeter, on amps which have suffered a red plate I have measured them anywhere between 320 Ohm (way low) to 880 Ohms (way high) plus the ones which just fall in half (way way high).

Cheers,
Ian
 
Hi Ian,

I have to admit I had not considered checking or replacing the screen resistors, so thank you, that's another worthwhile check.

I have a feeling once I replace the first pair of PSU capacitors, the B+ will be higher.

(Currently I measure 482Vdc, with 360Vac input to the diodes, even considering the double diode drop x 2, theres some more volts lost, I assume by the 100uF reservoir being past it's best)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.