Built a HoSo56 using 18w hoffman PT... voltages to high

[larstore]

Hello,

So i finished my first build. It is my version of the HoSo56, and it sounds good, but very gainy and a bit "thin".

1. After measuring voltages, i am pretty far off the schematic. Overall voltages are 50v+ above whats indicated.
I am guessing this is because the schematic indicates a 270 transformer, and i am using a 290? How would this affect the amp? (sound/reliabilty)


2. Using the 150 ohm cathode resistor, i seem to be running the EL84s at 16W... comments?


LT

[John]

Voltages are important with that circuit, IIRC, especially the voltages on the 5879 input tube. Yes, your PT is responsible for the higher voltage, just like mine was    (used the same trans). I used an old "real" 5Y3 to help drop the voltage some, and increased the dropping resistor values.


16W Pdiss is way too much for those '84s, they won't last long at that. And if you increase the cathode resistor your B+ will rise too.


Do a forum search for "my B+ is too high" and you'll come up with lots of threads.


As far as very gainy but thin, not sure what to say there. To lower gain you - very generally speaking - decrease the cathode caps in the preamp. But that will *generally* make it sound thinner. If you post a schematic, guys on here that know lots more than I do will be able to advise you  much better.

[larstore]

Thanks!

As far as i understand, there is not really a clean solution to drop B+ voltage without affecting other stuff...

I'll think i'll try to use a 5Y3 rectifier as well..  what dropping resistors values did you end up with?

LT

[ac427v]

Hi, I wanted a low voltage Bassman so I used a 5R4GYB instead of a 5AR4. It dropped the voltage from 450 to 405 and can handle a bigger power supply cap than a 5Y3. The only downside is that it is a big tube (1 1/2 X 3 1/4) Good luck with whatever plan you try!
--AC427

[sluckey]

Geezer shows B+ of 350V using a 270-0-270 PT with SS diodes. I know you can get 350V with that 290-0-290 PT if you use an EZ81. You'll get even lower if you use an NOS 5Y3. Either will work with Doug's PT.

What rectifier are you currently using?

[John]

I'll think i'll try to use a 5Y3 rectifier as well..  what dropping resistors values did you end up with?

I don't remember anymore. I never could get the amp to sound like I wanted, and at the time didn't know how to change the voicing of it. So, ended up dismantling it and making the one I use now.

[larstore]

What rectifier are you currently using?

I am using a sovtek ssr plug now. b+ is 375. My other plate voltages are off by 50+ V so i need to adjust dropping resistors all over i guess

[John]

Lars, if you pop in a tube rectifier like Sluckey said, you'll drop probably 30 volts easy. But 375 at the plate of the EL84s isn't *too* bad. It is higher than what they're happy with though. By use a tube rect. that should drop to 350'ish.


What voltage do you measure across your cathode resistor? (the 150 ohm)

[larstore]

Lars, if you pop in a tube rectifier like Sluckey said, you'll drop probably 30 volts easy. But 375 at the plate of the EL84s isn't *too* bad. It is higher than what they're happy with though. By use a tube rect. that should drop to 350'ish.

I just tried a 5y3, and that dropped the B+ all the way to 300 volts. EL84 at about 10W now.. Amp sounds much better. Will try to get a EZ81 and hope to get somewhere in the middle.

The main issue now is the lack of bass, and the tonestack doesnt make much sense to me. (little to no effect)
Turning the bass to 10, and mid treble close to 0 sounds the best.. however, i would still like more low-end...  Any suggestions? Schematic attached (By geezer, not me :))

What voltage do you measure across your cathode resistor? (the 150 ohm)

SS plug: 12.9V, 5Y3: 10.25V


edit: correct attachement now..

[tubenit]

The entrance cap into the LTPI,  change it from .0047 to .01

Change the tone stack slope resistor from 33k to 68k.   Change the bass cap from .02 to .047.

Look up (search) the "enhance cap" which is typically a 120p-390p cap across the plate resistor of the entrance triode of the LTPI. Try a 220p or 250p there.

That's where I would start.  IF needed you can change the .01 coupling caps to .015 between the LTPI and power tubes.

With respect, Tubenit

[larstore]

The entrance cap into the LTPI,  change it from .0047 to .01

On the post where i found the schematic and layout, this cap is actually .01 on the layout (made by you?) and 0.0047 on the schematic. Would you know why? Could you explain a tube amp noob what this does?


I was going to try a EZ81 tube or similar to get the voltages right, and maybe another speaker, but this is really helpful. Would you be able to give some thoughts behind the other suggestions?


Thanks!

LT

[tubenit]

I think it needs to be .01.  Probably just an error in drawing either the schematic or layout up.  It's simply a coupling cap.

Certainly the speaker you are using will make a HUGE difference.  I think the HoSo56 sounds good with an Emminence Cannabis Rex.

With respect, Tubenit

[larstore]

Certainly the speaker you are using will make a HUGE difference.  I think the HoSo56 sounds good with an Emminence Cannabis Rex.

I will try other speakers before anything else.

Regarding my B+ voltage, i need a voltage drop of 25V, not 75 which was the case with my 5Y3

Any suggestions for tube rectifier? According to http://www.300guitars.com/articles/rectifier-tube-voltage-drop-chart/ i may think a 5V4 could be suitable?


LT

[HotBluePlates]

Regarding my B+ voltage, i need a voltage drop of 25V, not 75 which was the case with my 5Y3

Why 25v and not the 75v you have now? Because it was on the schematic?

I don't think we're that precious with exact B+ voltages around here; 100-150v is a "big change" but for B+ 20-50-75v might not be that big a change (at least at the output tubes).

If it were me, I'd roll with what you have no.

[larstore]

Why 25v and not the 75v you have now? Because it was on the schematic?

I am not saying the schematic is perfect, but i've read alot about the importance of the B+ in this particular amp, and because its my first build, i was hoping to "go there" first, and start shaping the sound to my liking from there..

I will test another tube today to hear the difference.

[HotBluePlates]

I see what you mean.

Do you have any other 5v rectifier tubes handy? Like a 5V4, a GZ34 (or 5AR4), a 5U4, etc.? Each has a different amount of internal resistance and resulting voltage drop. It might not be a bad idea to have one of each (unless the price of buying them is a factor), because you might use them in the future and would have them available for swap and comparison as in this case.

[larstore]

I see what you mean.

Do you have any other 5v rectifier tubes handy? Like a 5V4, a GZ34 (or 5AR4), a 5U4, etc.? Each has a different amount of internal resistance and resulting voltage drop. It might not be a bad idea to have one of each (unless the price of buying them is a factor), because you might use them in the future and would have them available for swap and comparison as in this case.

At the moment i don't. However, i am getting a 5AR4 to test today

[tubenit]

Do you have any other 5v rectifier tubes handy? Like a 5V4, a GZ34 (or 5AR4), a 5U4, etc.? Each has a different amount of internal resistance and resulting voltage drop. It might not be a bad idea to have one of each

That's exactly what I do!  Like in my D'Mars,  I use a 5Y3GT for 6K6 or 6V6,  a 5V4 for 6V6 or 5881 and so on......

With respect, Tubenit

[terminalgs]

The main issue now is the lack of bass, and the tonestack doesnt make much sense to me. (little to no effect)
Turning the bass to 10, and mid treble close to 0 sounds the best.. however, i would still like more low-end...  Any suggestions? Schematic attached (By geezer, not me :))

even before the slope resistor and the rest of the tone-stack (T/S), or the PI is the parallel 470K-R and the .02uf-C that connects the CF to the T/S.  you can find online calculators that will tell you the impedance and phase-shift of that R||C for different frequencies.  for low frequencies, the impedance is high,  and for high freq., impedance is low (relatively speaking).   This impedance is not as much of a concern for the relationship of  preceding CF's output impedance  as it is a factor in your T/S.


Pretend for a moment that the rest of the T/S has the same impedance for any frequency (it obviously does not, but lets pretend).  Think of it as a static resistor to ground rather than the network of R's and C's.   the total attenuation effect of the network made up by the 470K-R||.02uf-C and the pretend resistor results in a filter where lower frequencies are attenuated more than higher frequencies. Especially, if you compare that to the same circuit where a straight piece of wire exists in place of that 470K -R||.02uf-C.


So that's what I'd try, tack in a jumper on parallel with the 470K-R||.02uf-C and see what happens.


Of course, the T/S isn't a simple resistor to ground, but whatever impedance the T/S presents for whatever frequency,  the 470K-R||.02uf-C pair will react to it in a similar way.


that's my .02c. on where to start.


Of course, for any coupling capacitor or series capacitor, you can find the impedance to ground following that capacitor and use an online calculator to examine the bass-roll-off effect of that particular cap/resistor combo.  When you change a coupling cap from .001 to .01,  the cap by itself doesn't necessarily effect different frequencies any differently, it is the impedance of the cap AND the impedance to ground immediately following that cap that determines what frequencies "pass" with less or more attenuation than other frequencies.  you can make these calculations for every capacitor the audio signal "passes" through...   see http://en.wikipedia.org/wiki/High-pass_filter

[larstore]

Do you have any other 5v rectifier tubes handy? Like a 5V4, a GZ34 (or 5AR4), a 5U4, etc.?

Tried a GZ34. That dropped the voltage to about 360V.
By changing a dropping resistor also, all my voltages are 5-10 volts above schematics. My EL84s still running a bit hot at 14W or so...

The entrance cap into the LTPI,  change it from .0047 to .01

Tried. Not sure if it affected my problem, but i'll leave it for now

.
Change the tone stack slope resistor from 33k to 68k.   Change the bass cap from .02 to .047.

Tried this, and it did not make too much of a difference (will test this some more since it was late, and i couldn't turn up the volume)
However, and am not sure if the tone stack is my main problem. I tried to bypass it alltogether (jumpered directly from the CF directly to the PI), and it still does not give me what i want.

So that's what I'd try, tack in a jumper on parallel with the 470K-R||.02uf-C and see what happens.

This definately helped some. What does this RC do?

[terminalgs]

If your voltages are 5-10V over, I'd say you are in good shape for now on the voltage.

So that's what I'd try, tack in a jumper on parallel with the 470K-R||.02uf-C and see what happens.

This definately helped some. What does this RC do?

higher freq. see that pair as low impedance (pass AS-IS), lower freq. see that pair as high impedance (pass attenuated).

however, i would still like more low-end...  Any suggestions?

Back to this ^^^.

At pretty much every stage of this circuit, from input jack to power tubes, this amp looks to be built for high-end freq. and not low end.


Look at the gain stages and the cathode by-pass capacitors.  Compare them with the bypass capacitors in a '59 bassman.


Even better, compare them with this amp's original inspiration: the AC15.  An AC15's final cathode by-pass cap is a 50uf.  The same AC15's EF86's by-pass cap is a 25uf cap, and although that same AC15 doesn't have a topboost (TB) gain+CF between the EF86 and the PI, the original AC30 TB gain stage's bypass cap was a 25uf as well (if it was fitted to the circuit at all, I believe it was omitted completely in some TB's...  like your relay would do).


Your amp has the following: 2.2uf on the pentode, 4.7uf on the TB, and finally 8uf on the el84's.  You can read about the effect of the bypass cap in many places (including valvewizard's gain stage chapter) but in real general terms,  on preamp stages, anything less than 25uf will create a gain stage that provides more gain for higher frequencies, and less gain for lower frequencies.   On the el84's it is similar, only 50uf-100uf is usually the preferred value to provide the same gain to all frequencies.


(just to pick on the pentode's cap:  with a 470R and a 2.2uf, the low freq. roll off is about 150Hz.  that means anything below that will not have the same gain as freq. above it.  that includes the open string notes on your 5th and 6th strings..  you'll need to bump it to 4uf or 5uf to include those frequencies


Also, look at that wikipedia link about caps and resistors and how they form high-pass filters.   Every coupling cap in the amp can be examined for the high pass effect.   One secret about the long-tail-phase-inverter (LTPI) that this amp has is the impedance of this LTPI (following that .0047) is 2Mohm (google 'long tail phase inverter impedance'). what this means is the .0047 cap and conjunction with the 2M impedance of the LTPI doesn't create any significant bass-roll-off for guitar.  Switching this .0047 to a bigger cap won't really do anything, in fact, it can cause the amp to thump, or pass some LFO (say from a trem/vibrato oscillator--doesn't apply here...)


like mentioned by others, speaker's can be dark, or bright.  Try the amp with other speakers you have in other amps with a patch cord and alligator clips.

[larstore]

Will experiment more this weekend, just a few quick questions:

higher freq. see that pair as low impedance (pass AS-IS), lower freq. see that pair as high impedance (pass attenuated).

Should i keep it bypassed, or look into this and change the values?

At pretty much every stage of this circuit, from input jack to power tubes, this amp looks to be built for high-end freq. and not low end

Listening to the "original" soho video here, 65Amps Soho w/ Peter Stroud
i don't get that "depth" at all.. it leads me to think that i have something wrong.... but then again, i haven't tried it with a LP

like mentioned by others, speaker's can be dark, or bright.  Try the amp with other speakers you have in other amps with a patch cord and alligator clips.

Tried another speaker yesterday. Sounds pretty much the same..


It is of course possible that this is the way it should sound. I cannot find any sound clips from other builds..

[Willabe]

i haven't tried it with a LP

What guitar(s) are you playing though it?


            Brad   

[DummyLoad]

change the two .01uF between power tubes and long tail phase inverter outputs to .022uF.


as-is at .01uF, the roll off is ~63Hz.


--pete

[terminalgs]

I think we are all making comments about getting more low end out of the amp while assuming it is built correctly and there no problems with the amp itself (I know I've been assuming the amp was properly put together).


So, that said:  Check and double check your solder joints on your filter caps, both on the B+ side and the ground side.  If you have a filter cap (any of those under "E", "D", "C", or "B" ,,  but mostly "E" or "D") that isn't functioning (bad ground perhaps?) you may not have good isolation between stages.  Especially with a CF, where the audio signal from the plate of the CF is directly on the B+.  the large cap looks like a ground to all audio signal --if functional-- but if it isn't functional, the signal can bleed to other stages and mix on other plates, sometimes at out of phase relationships, canceling some or all of a signal.  Depending on the other caps and resistors in the B+ rail, different frequencies could be more pronounced and available for signal canceling.


Kind of a long shot, but its not a hard thing to test and examine.

[larstore]

What guitar(s) are you playing though it?

A strat with humbucker, and a fender jazzmaster so far

change the two .01uF between power tubes and long tail phase inverter outputs to .022uF.
as-is at .01uF, the roll of is ~63Hz.

Hmm, i might try that, but i still supect something else..i don't think this is the way the "original" Hoso65 sound + the sound clips i've heard of the soho sound much fuller/richer

My 5879 tube was bought used online.. i have ordered a NOS to test, just to verify that nothings wrong there.
Any other simple tests i can try? like bypassing certain parts to "narrow" down where the "problem" might be? As it is now, its pretty much "useless" without bass at 10.. and even there its still not where i want it.

[tubenit]

i don't think this is the way the "original" Hoso65 sound + the sound clips i've heard of the soho sound much fuller/richer

I built the HoSo56 originally to use EL84's and then switched to using just the pentode of 6BM8's.  I played it that way for a couple of yrs with a band before I sold it.

Here is a sound clip of my HoSo56  (56T) so maybe that will give you some reference point regarding the tone. I did not find that amp too trebly.  So, maybe you have something wired up wrong?   (Please excuse the mediocre and repetitious playing)

http://www.soundclick.com/player/single_player.cfm?songid=7988948&q=hi&newref=1

IF you'd like a better sounding soundclip for comparison,  I can email you one that is "City Lights" by Journey that I think recorded better.

With respect, Tubenit

[John]

As it is now, its pretty much "useless" without bass at 10.. and even there its still not where i want it.

Just now catching up. This amp has plenty of bass, as built to schematic; I always found it a dark, lush sounding amp. I'm inclined to agree with Tubenit, there's probably something simple mis-wired. You find that and you'll find your bass.

[tubenit]

I wished I had access to Geezer's HoSo56 sound clips because they are incredibly great sounding.

Per Geezer's comment on the "pseudo channel switching"  HoSo56

Wow!  This amp is more "Dumble" sounding than any of the "real" D-style amps I've built, plus it's a very simple circuit.

And from what I remember about the sound clip he did that's pretty accurate.

I've attached a Hoffman style layout.  So, check your wiring compared to the attached layout.

This is a sound clip that was done using a HoSo56 by a Hoffman forum member  Craig Barnett:

http://mixcraftlive.com/index.php?t=popup_player&mode=song_hifi&band_id=532&song_id=1799

With respect, Tubenit

[stevehoover]

I had the same voltage issue on a similar build, B+ was 340 on the output plates with an 5AR4 or a GZ34S

measured from the plate to the cathode.


I found that using aSovtek 5y3 (the one that the experts say isn't really a 5y3) brought me down to the 320 range where I wanted to be.


This tube seems to lie in between a 5AR4/GZ34!and a real 5Y3 Voltage drop wise.
It's also a slow turn on like the GZ34 so it doesn't slam the filter caps as hard when you power up.
It's an option some have forgotten about but a decent rectifier.

[Ed_Chambley]

I wished I had access to Geezer's HoSo56 sound clips because they are incredibly great sounding.

Per Geezer's comment on the "pseudo channel switching"  HoSo56

Wow!  This amp is more "Dumble" sounding than any of the "real" D-style amps I've built, plus it's a very simple circuit.

And from what I remember about the sound clip he did that's pretty accurate.

I've attached a Hoffman style layout.  So, check your wiring compared to the attached layout.

This is a sound clip that was done using a HoSo56 by a Hoffman forum member  Craig Barnett:

http://mixcraftlive.com/index.php?t=popup_player&mode=song_hifi&band_id=532&song_id=1799

With respect, Tubenit

Just watching, but I do have a question.  Why the choice of the Gibson pentode over the ef86?  I know this is not the bass problem, but those tubes sound much different.

The 65 SOHO is a ef86 amp.  I have played the SOHO and it is really a nice amp, not just another version of the 1974 Marshall nor does it sound like a early Vox.

Also, looking at the original version 1 of this amp it had a 1m Master pot on the screens.  How well did that work?

[tubenit]

Why the choice of the Gibson pentode over the ef86?

It's an incredibly great sounding tube.  I've had NO problems with microphonics which seem to plague many EF86's and there is less gain.

And it's reasonably priced.  I've never had an interest in building with an EF86.

with respect, Tubenit

[larstore]

Just now catching up. This amp has plenty of bass, as built to schematic; I always found it a dark, lush sounding amp. I'm inclined to agree with Tubenit, there's probably something simple mis-wired. You find that and you'll find your bass.

I have rechecked my wiring several times, and i cannot find anything incorrect. There might of course be something wrong with the components,
As previously mentioned, i have tried to bypass the tone stack. How should i expect that to sound?
And, do i have any other quick tests i should try to narrow down the problem?

LT

[John]

Lift the ground connection on your tone stack, that pretty much disables it and lets the pure signal through. Double check the cathode bypass caps, make sure you soldered the end that goes to ground properly- no "cold joint". You probably already did all this, but it's what I'd do first.


Trouble shooting is my weakness, although I'm (thankfully, and by necessity lol) better than I used to be. When I first build the HoSo, it had very very low volume. I spent probably a total of 6 hours over the course of several days tracing and re-tracing wires, highlighting the schematic each time, and FINALLY I saw what I should have seen the first time. I forget exactly what it was, but something to do with a filter cap connected where it shouldn't have been. 

[larstore]

So, i have been working my way through the schematic and layout. I have been trying to calculate the different frequencies for the caps throughout the amp. As far as i can see, there is nothing wrong with the preamp or PI that should make the amp sound thin and bassless.

However, I have found a couple of inconsitenies between the layout and schematic.

1. The PI input coupling cap is .047 in the schematic and .01 in the layout. As already mentioned by Tubenit, this should be .01. Didn't make much difference when i changed it though.
2. The EL84 screen resistors is 100 ohm to each tube, in series with a 1K resistor on the schematic, the layout lists just 1K to each tube. Should not make a bit difference i guess
3. The EL84 cathode bypass cap is 8uF in the schematic, and 100uF in the layout. According to my investigation, this cap is normally pretty large to avoid bass loss? The AC15 uses 220uF. I have located other HoSo56 schematics, and they also lists this as 8uF?

[John]

Something easy to try is either tack in or use alligator clips to parallel a 47uf cap across the 8uf that's already there. I doubt it'll make much difference by itself, but you never know.

[larstore]

Success!

After staring at the circuit for a while, i noticed a resistor under some wires with no connection... That happened to be one of the plate resistors for the PI.
In other words, one of the PI tubes had no plate voltage at all. In addition to this, i found a miswired preamp grid.
Got pretty late yesterday, so i need to test with some volume today, but i am pretty sure this solves most of my issues.

I have no idea how i managed to overlook these mistakes earlier, but atleast i know alot about the circuit now ;)

Thanks for all the ideas, i guess i will start new threads when i start "shaping" the sound to my liking ;)

LT

[Willabe]

 

Take a look at this link that Doug posted on using a highlighter on a schematic to make sure the amp is wired up correctly and every solder connection is soldered up. It works great.

http://el34world.com/Forum/index.php?topic=17701.msg178630#msg178630


                 Brad