newb express 6v6 vs. el34 question

[pinkmarkos]

hey guys- i hope some TW guys are browsing over here... i just got my first ever build off the ground, an express. BTW- it f***in rocks! my question is this- when using 6v6's biased at 21-22ma with 411VDC on the plate to give about 18 watts total- my PI grid voltages are about half of what they were with el34s (which i had biased at 43-44ma). this is normal right? the plate and cathode voltages are consistent with what the guide says i should have, and all my B+ voltages are good. bias voltage is -27VDC. hopefully someone smarter/more experienced than me can explain this to me. i just want to make sure everything in operating as it should. Thanks!

[Phil_S]

Your grid voltages should be very small numbers. If plate and cathode are checking out, I would guess everything is fine. Please post plate, grid, and cathode comparatives for the PI so we can actually see what you are talking about. It might get you a more definitive answer.

I think you said it sounds fine. That's a good sign.

[pinkmarkos]

hello- i just went to work so i will post those voltages when i get home. i will post the voltages i am getting vs. what the voltage guide states i should be getting at the PI. two things- 1. this is my first build, and 2. i haven't heard another express so i don't really have anything to compare it to soundwise except the geetarpicker clips on you-tube (and it doesn't sound nearly that good) . sounds good to me up until about 12 or 1 o'clock on the volume. good in this case means nice power tube break up a la plexi. the clean sound is also very nice and chimey to me. i play a les paul FYI. after 1 o'clock or so, it seems to get slightly unstable (oscillating). tone controls and prescence all work well. thanks for looking- MP

[pinkmarkos]

hello- here are the PI voltages I am getting and the expected values-
V3a- pin 1/plate 192 (190 expected)
pin 2/grid- 13 (28 expected)
pin 3/cathode- 38 (42 expected)
V3b- pin 6/plate 196 (193 expected)
pin 7/grid- 12 (27 expected)
pin 8 cathode- 38 (42 expected)
These are the same regardless of biasing for 6v6's or el34's. The B+3 voltage that feeds the plates of v3 is 313 (317 expected), and my B+4 voltage that feeds the plate of v2 and the grid of v3b is 295 (298 expected). i suspect maybe a bad capacitor, C9 on the express schematic. how do you check for capacitor function with the DMM? I would appreciate anyone who could try to help, so I can understand for the future. thanks!

[Phil_S]

There is always going to be some difference from expected values. Line voltage and the fact that parts are not military spec, etc, are the main factors. Your PI results don't seem all that out of line to me. You get the uneven plate voltage because the plate load resistors are different values.

Your grid voltages are not so far off from mine. I built mine to use 6AQ5's instead of 6V6's, so I designed the final stage to run at lower voltage, but the preamp section is pretty much stock.

My PI has about 12V on the cathodes, about 9V on the grids, and plates are 187/205. The higher voltages you are using in the final/power section will give results different from mine.

Why would you suspect a bad cap in C9? Your voltage readings suggest otherwise. The circuit is symmetrical in it's feed to both sides of the PI, so you should expect nearly identical grid voltages. Your grid voltages are 12/13. If you had a bad cap, your spread between these two values would be significant.

If you are concerned about low grid voltages, you might want to check and double check the parts between the B+ supply and the grids to be sure they are all in spec. Take a good look at R10-R11-R12-R13. It is so very easy to insert a resistor of the wrong value, like 100K for 10K or 470K for 470 ohm. To properly check them with your meter, it will be necessary to have one end of each resistor out of the circuit. You can probably get all of them but the 10K by pulling the tube. You might need to desolder one end of the 10K. Any funky readings dictate that you lift one end out of circuit by desoldering before you jump to any conclusions.

Moving on, regarding your bias voltage, are you saying that your maximum negative voltage is -27v? You should be able to get below -33v. To modify the bias range, change the value of the dropping resistor R29. Use a lower value (180K, 150K) if your bottom is only -27v. Changing your bias voltage is likely to have an impact on other readings, including the PI. See below, -27v might be just fine.

As far as bias voltage and plate voltage, you can't expect same result for 6V6 and EL34. Your 6V6's are running at about 18W and EL34's at about 36W. These are reasonable results.

This is your first build? Balls to the wall! This is not your typical first project. This is a difficult amp to get right. My hat is off to you that you have so few things that really need work. I am not surprised at all that you've got problems develop as you crank the volume. There is a good chance that the problems are lead dress related. Lead dress, the length and placement of wires, is a bigger factor in getting good results in this type of amp than in many others. If you do not have a scope to show you exactly where the problems are developing, then you must go by trial and error. Using a chopstick or other non-conductive (no pencils), wood or plastic, probe and move various wires while the amp is powered on to see if you can improve on what you've done. Also, you might tap on components, and you might find that shifting a cap or resistor a bit is the thing to do.

Good luck, and don't give up on making the amp the best it can be. Enjoy the journey as much as the result.

[pinkmarkos]

phil s- thanks a lot for explaining this to me, i feel better knowing that my voltages are within reason. my max bias voltage is -29.5 VDC measured at pin 5 on the el34's. my bias was set at -27VDC (40ma) to give about 32 watts total power. i think i will take your advice and switch out R29 to something lower.
thanks for the vote of confidence with regards to the build itself. you are right in saying that this is not the typical first build, as i have discovered . with this amp, as many people have said before me, everything matters. i have a feeling that tube selection will really help me get to the tone in my head with this amp. i have some xf2 mullards on the way now that i am excited to try. also, this amp will exaggerate any microphonics at all, esp on v1. i use EI's for tone, but they are really microphonic, so i need to find something else. how do the 6aq5's sound? different pinout and plate voltages I would imagine, as you alluded to. BTW- do you have a liverpool as well? i would like to try that for my next build once i get this one squared away. thanks again for your advice.

[Phil_S]

For testing purposes, probably the best tubes to use are Sovtek 12AX7's. They are known to be quiet and will take abuse, but they are not known for tone. For power tube testing, I think I'd look at a pair of JJ 6V6 which I've heard will stand up to obscene plate voltages.

So, you've taken the plunge for Mullards. Well, I hope you like them in this amp. IMO, this is a high gain amp and you may find that it does not benefit significantly from the Mullards, but let your ears decide. My experience with EH 12AX7 in this amp is that I didn't like them.

The 6AQ5 is just a 6V6 in a small bottle. Honestly, I've got only the same frame of reference that you do -- the same youtube clips of the same players. I like the sound just fine. I have a collection of about 50 6AQ5's and related substitutes, so I had lots of choices. I settled on one black plate Sylvania and one gray plate Zenith branded probably Sylvania. These are pretty tightly matched and current draw suggests these are new or almost new.

I have several amps that I've built, but the 'wreck is the only KF inspired design. Here's my latest, a bit of a Franken-amp:
http://home.comcast.net/~psymonds/6SN7amp.htm
If you click on the sound clips, you will be suffering through my terrible guitar playing, so you've been warned. Also, the sound samples are made with primitive equipment and the clipping you hear is the recording, not the amp, which is pretty much pure clean tone.

[pinkmarkos]

hey phil_s- i use the jj 6v6's as my 6v6 choice because they will handle up to 450 volts on the plate. i have some other GE/RCA 6v6gt's that i am not sure can handle the 410 volts i have on the plate so i don't use them in this amp. my theory on the mullards is that the express, from what i've read, is designed to get power tube distortion first, then the PI, then the pre-amp- so better sound can be obtained from using better power tubes. the jj's actually sound very nice in this amp, IMO. last question- when you suggested that i check the grid supply resistors, do i need only to remove the PI tube, or all the tubes? i ask because i can get readings with the tubes in that seem accurate. thanks again for your advice.

[Phil_S]

Whether you get an accurate reading depends on how the other parts of the circuit interact. Pulling the tube leaves one end of the connection with nothing there, effectively lifting it out of the circuit. Sometimes, just having the power off accomplishes the same thing.

If you are satisfied the parts are correct, I think you need not do anything else.

You might swap in a different tube for the PI and see if the voltages change. It could be the tube and not the parts.

Anyway, it seems to me things are OK. Don't sweat these details if the amp sounds good and you're not letting any smoke out.