7189 questions

[luthierwnc]

Hi Guys,

It's been a while.

I am looking at a Teisco Checkmate 18 for a friend. I haven't worked with 7189's before and this bad boy seems a bit out of spec. The schematic is at schematicheaven in the bargain bin section and it looks very accurate.

My problem is the power supply. At full juice, it is showing about 350 VDC on the plates. The screen node is downstream on a 5k series resistor. The screens are showing around 230 VDC. Each tube is pulling about 95ma (shunt test). One of the problems my friend was having was that the first power tube was red-plating. It was pulling 130ma and the other was pulling 4. Removing the tremolo wire from the power tube grid leak junction took care of that for now (keep your eyes peeled for subsequent posts). The power is very low and it sounds like, well, not good.

My research suggests that screen should be pretty close to the plate voltage. It is a first cousin to an EL84 and they usually don't have anything like 5k between the nodes. Voxes are just a choke and a 100R resistor. That also seems like a lot of amperage with the screen and plate so far apart.

If you have worked with this tube before or have some ideas, I'd appreciate hearing them.

Thanks, Skip

[Robert M. Martinelli]

Hi,
7189s are basically mil-spec beefed-up versions of 6BQ5s/EL84s,
( BTW a 130 Ohm cathode resistor is exactly the recommended one for class AB1 operation with EL84s ) but, if you have had 130 mAmps flowing through one or both the tubes, I strongly suspect both your 7189s have definitely gone south - the tube drawing 4 mAmps only surely has ( .13 Amps @340 V is overkill ).

The "bad sounding" symptom is almost certainly due to the output section imbalance, as one of the two halves of the waveform is missing.

A 120 VDC drop across the 5 K resistor means 24 mAmps are drawn by the 7189s screens and the preamp/PI tubes, my ( educated ) guess is that the preamp tubes draw some 4-5 mAmps, this means the screens draw 10 mAmps each, which is a bit high IMHO.

I'd first check all the passive components in the power amp area, especially the cathode resistor ( if the bias voltage is too low the tubes are likely to "redplate" and die - some 12 VDC quiescent on the cathode would be OK ) - Also perform a DC resistance check on two halves of the OT primary to check' em for symmetry, then try to change the tubes and take new readings in both quiescent and working conditions.

Hope this helps

Best regards

Bob

[black_labb]

7,189... thats alot of questions!!!

rob, he said in the post that by removing the tremolo connection he was able to get a better balanced output. i dont have much tips to help though. the 5k resistor does seem different, but not neccesarily bad. as rob said the screens are probably pulling more than they should so rebiasing may be in order. i dont know where i would go after that. maybe a bit more of a poke around at voltages and such to see wether anything is causing a bad sound due to the voltage on it.

could it be that there was meant to be higher voltage on the plates than common and the 5k was to drop the screens to more manageable level's? being called a checkmate 18 they may have expected to put out 18w which i really doubt with screens significantly lower than an ac15 (i am assuming there are a single pair of output tubes).

i just looked at the schematic, and with 5k off the plates and then 10k followed by 30k to get to the plates of the firstgain stage i doubt that they wanted 230v on the screens followed by alot of dropping resistors to feed the first stages plates. 230 wouldnt be too high a supply voltage for the preamp. i'd check the rectifier incase they wanted more voltage (but 350v is probably not too far off) and for something drawing more current than it should. leaky filter cap?

i assume that the sound issue is to do with low voltage on the preamp and PI, and possibly not enough gain/headroom in the output section due to the low voltage on the screens.

i know very little about trem systems. maybe a leaky cap is in the trem system and is causing the trem to not oscilate and is what was causing the bias issue/redplating before disconnecting. it doesnt make complete sense, but i'm sure someone else has a reasonable explanation

[luthierwnc]

Thanks guys,

This is one of those cases that will probably take an equal amount of theory and fiddling. The owner (a decent amp tech in his own right) left everything bone stock and was getting the runaway current in the V4 tube. We had been emailing back and forth for a couple months about what it could be and I finally told him to bring the thing to lunch one day.

In the back-and-forth, I'd wondered about the grid leak resistors and the way they only ground AC through that .05 cap. Since the usual cathode-biased EL84 amp grounds everything at that point, I thought some DC was getting on the grids. At any rate, removing the speed wire from the junction got the amp to a more normal operating situation so I'll leave it off until I can make the thing sound good without it.

I don't have a spare set of power tubes so I'll swap in a known-good pair of EL84's and see if the tubes are just funky (have to move one of the screen wires). Robert: you wouldn't happen to know the screen volts on a VOX or Liverpool by any chance?

I'll stay in touch, Skip

[luthierwnc]

further developments

I swapped in the EL84's and got the same thing. With the grid leak resistors going through the .05 cap the current is very high but not that much different between the tubes. With the grid leak resistors grounded, the first power tube draws 145ma and the other hardly any. The next thing I'll try is shotgunning the PI components to see if the bottom half is starving the second power tube. sh

[luthierwnc]

fixed -- mostly

Hi Guys;

It was a bad .05 coupling cap coming off the V2 plate.

Now it is putting out 40 ma on the top tube and 51 ma on the other. Not much volume but it may always be that way. One weird thing, I ran a 1k sine wave into the amp and it never distorted. With the reverb and trem off, the wave form was identical (except for the voltage) from the input jack to the speaker jack -- even dimed. I've never seen that before.

Volts are 408 on the plates, 364 on the screens and 13 on the cathode. The first three preamp plates are around 140 and it goes to 260 on the plate of the inverter.

I'll monkey with getting the draw more even tomorrow.

Thanks again for looking and good luck with your projects.

Skip

[black_labb]

the fact that it is quiet and never breaks up suggest that there could be a fair bit more gain in the amp. what kind of amplitude did you have on the sine wave?

edit:at the imput i mean

[Robert M. Martinelli]

Hi Guys;

It was a bad .05 coupling cap coming off the V2 plate.

Now it is putting out 40 ma on the top tube and 51 ma on the other. Not much volume but it may always be that way. One weird thing, I ran a 1k sine wave into the amp and it never distorted. With the reverb and trem off, the wave form was identical (except for the voltage) from the input jack to the speaker jack -- even dimed. I've never seen that before.

Volts are 408 on the plates, 364 on the screens and 13 on the cathode. The first three preamp plates are around 140 and it goes to 260 on the plate of the inverter.

I'll monkey with getting the draw more even tomorrow.

Thanks again for looking and good luck with your projects.

Skip

Hi Skip, ( hi bl ! )
Glad you're on the right track.....
( wow, i was only 1 V away with my cathode voltage guess..... )

As to the amp not breaking up, it could be either a matter of low preamp/PI gain ( check each stage's gain ) or simply the signal you injected had too low an amplitude....or.....read the last paragraph....

About the 40 and 51 mAmps ( quiescent ), apart from the imbalance, I think they're both too high, even for 7189s ( 40 mAmps @ (408-13)VDC = 15,8W /// 51 mAmps @ (408-13)VDC = 20,1 W ) let alone EL84s. I think that, keeping the plate voltage the same, some 32-33 mAmps would be advisable to keep the tubes from failing prematurely, so I wonder if it's advisable to try different cathode resistors ( 150 or 180 Ohms ) to be on the safe side. ( Your amp has been designed when the mains voltage was 110 VAC, now AFAIK in some cases it can be over 120 VAC, causing the +B voltages to be some 10% higher thus stressing the tubes/components, hence the need to "cool things down" ).

BTW, the higher voltages could have the side effect to expand the amp's headroom, and this could explain why the amp's not distorting....with lower voltages it would probably "break up" at a certain point.

Cheers

Bob

[luthierwnc]

I replaced the 30k supply resistor with 10k and it took the preamp voltages up to around 150VDC. I agree that the amp could use more gain but that will be Dave's project. Even diming the bass control has very little effect and the treble has to be close to 10 to get much out of the amp.

Me, I'd pull the existing stack and put in a blackface set-up. Either that or put in a push-pull tone stack bypass. These little amps don't have enough signal to throw that much away.

Thanks and I'll see you here again, sh