Is hooking up a battery to preamp a good indication of what dc power will do?

[frank57]

Well I hooked up a battery to the heaters on my Hiwatt and it seems to solve 95 percent of the hum buzz,bad layout and every other problem these amps have.The buzz is totally gone.
Unless I did something wrong there.
I left the center tap from the transformer on the ground bolt for the fuse, where it was.
The tubes are not lighting though but I get a guitar signal and the amp is working.
411 on the power tubes, bias around -17.34.
Does voltage go up with some kind of dc power?
Now is it a good indicator of what elevating the heaters or a dc power board would do?

[kagliostro]

The tubes are not lighting though but I get a guitar signal and the amp is working.

How many volts there are on the filament with the battery supply ?

if the tubes don't light as usual may be you lost a part of the charge of the battery and you haven't the 6.3v (or the 6v)

as you have a PBC I think it will be better to try a DC filament PS instead than an elevaded heater PS

with an elevated heater PS may be you are fine also in a PCB circuit, but as it wasn't planned to do, I'll prefer don't do it

Kagliostro

[frank57]

I'm not sure about the voltages on the preamp tube heaters on pins 4and 5.
I don't have it hooked up right now though.
But everything was definitely working as it should but not lighting up.

Battery is brand new.I don't know what should happen when you hook up a battery to do this test.
The guitar signal was as loud as it should be.So that was okay.The voltages on the preamp tube
plates were ok too.The hum and buzz were pretty much gone.
If the test is a good indicator of what dc power could do then it might be
 the way of solving it short of rebuilding it all.(you might still need dc power even with a rebuild)
If the battery wasn't supplying enough power would the amp still work?


I'm not quite sure how to do the dc power board though.

[kagliostro]

I'm not quite sure how to do the dc power board though.

Many DC heater PS are dedicated only to V1, other give the voltage to all the tubes of the preamp but I think I've never seen Power Tubes with DC on heaters, also if that can be done, but you need more current.

to give power to the heaters you have 3 different choices

.give a DC voltage
(simply rectify and filter with a couple of medium sized (in uf) caps, say 2 x 4700uf (a couple of cap is better than one big cap because they afford better the ripple than the same value in only one cap, or you can use only one very good quality electrolytic cap)

.give a stabilized DC voltage

.give a stabilized DC current

for a stabilized DC voltage you must have more AC voltage, say 12v for a good, real 6.3v stabilized DC voltage

if you do a Google search you can find documentation

http://diyaudioprojects.com/Technical/Voltage-Regulator/

http://www.pastisch.se/faktiskt/HeaterSupplies.pdf

for stabilized supply an LM317 is good for 1A consumption circuit

for higher consumption you can use one LM338

Kagliostro

p.s.: can you post the schematic of the PS of your amp ?

[frank57]

The power supply is something like this.
I think the original setup is very similar to a typical Hiwatt power supply more or less.
The board changed things a little but I'm not sure how to draw it in there.
The original amp did not have a standby switch.
It's a Hiwatt style circuit sort of like an overdrive lead.
The original tonestack is similar to  the tone stack I think from a 60's Hiwatt.
They did use something like that on the odd Hiwatt here and there.

[kagliostro]

I've seen the pdf file, but where is the heater power supply ?

If we can see it maybe we can say something to help you

Kagliostro

[frank57]

Well that's part of the problem.
There is no official schematic for the amp.
You can see the heaters on the tube pcb in the board layout.
But is there anything I can do to check that green wire from the power transformer?
Maybe it's a lousy center tap?

[kagliostro]

Try to draw the heater schematic.

There is a 0-6.3v or a 3.15v-3.15v transformer ?

there is a virtual ground ? if not have you considered a Humdinger ?

try do disconnect the tubes from the PCB and connect directly it, with a strictly twisted wire, to the transformer

I'm just trying to help you but I'm not so experienced, hope a more experienced member of the forum will give you help

Kagliostro

[frank57]

Does this help at all?
There is no virtual center tap.
The only possibility is that green wire from the power transformer.
Maybe the center tap in not very good which might explain the high voltages?

[kagliostro]

May be the better way for your PCB circuit is to have a DC heater supply

as you have power tube near preamp tube try also to give DC to the power tubes

if you don't want to use an IC you can try the Marshall TSL100 heather circuit

or that used for V1 in Mesa Boogie DualRectifier

if you want an IC look to the Mesa Boogie MarkV using LM2940 for V1 - V2 - V6

( http://docs-europe.electrocomponents.com/webdocs/06e8/0900766b806e8405.pdf )

remember that (IMHO) you preferably must supply also power tubes with DC, so take that into account to calculate the current necessary and the number of IC if you decide to use it

Kagliostro

[ThermionicEngine]

Greetings guys,

Interesting thread.  Couple of thoughts.  If the filament is not lighting, your tube is not delivering designed gain or characteristics.  This condition likely stops a lot of hum because the gain is likely to be low.  Kinda like running the amp on a current limiter.  You might get "plate voltage OK" at idle but I think the cathode current is not going to be OK and the voltage should sag consequently.  Therefore your battery hookup is not going to tell you if a regulated full power DC will solve the issue.   A drawing of what you have or a picture would help determine the original hum issue. Other questions- are you sure it is heater hum?  What frequency is the hum? If there is no center tap from the heater supply (usually a green striped wire) you can easily make an artificial one with a 100 Ohm resistor to ground on each heater wire.  Another tip would be to make certain the heater wires observe the same polarity for all the pre-amp tubes.  I presume this amp is a push pull power amp?  If so, there is no practical use to putting DC heater supply on the output tubes as they usually phase cancel that hum.  With all due respect, I would re-examine the hum issue to be certain you rule out other causes beside filament supply.  Whenever I get loud hum I start my T-shoot at the rectifier/filter circuit.  BTW, if you get hum at 120 Hz it is almost surely a filter cap issue.  Bad layout is not likely to be corrected with DC heaters.  Unshielded long grid runs will hum like crazy regardless of filament supply.  Ultra high gain amps do benefit from a DC filament at V1 and possibly V2, but lots of great amps have no DC filament supply and sound great.  If you can get some pics or drawings that would be great.  A chart of tubes and voltages at each pin would also be awsome.  Regards, Larry

[frank57]

The cathode on v2b seems to be a buzz source.If you remove it buzz goes down hum up.
Take out v2b and into a low gain situation and the hum buzz is very small and negligible.
All of these amps have this problem of hum and buzz.The hiwatt fix was 2 15k cathode resistors instead of 1.5k.It does not work.

It may simply be a  bad layout or a botched power transformer run maybe.Grid wires are shielded.
It's not something like that.It doesn't appear to be coming in on the ground of the pcb.The tracks were cut and though helpful didn't do much same with twisted heaters etc.Doubtful if it's a filter cap problem, it was checked out by some repair people who didn't know what to do other than bias it at -25 I think it was.

I f the tubes are not delivering the gain they should would the distortion go down?In this case everything was as it should be. It was still a high gain amp with guitar plugged in.I'm not sure what pin 4 on the el84 is doing but I took a shot at drawing the heater connections.
Only thing that's really worked is the battery.
Any tests I could do on that center tap?

Quick test:
I measured the resistance (with amp off )
from the disconnected heater wires to the earth bolt where the center tap supposedly is.
One yellow wire gives me 0 ohms.
Other one gives me 37k.(I guess this one is going through the diode?)

[ThermionicEngine]

Hi Frank,

Plese check the schematic against what you see- it appears to be an earlier version.  On your heater diagram, are all preamp tubes connected the same as V1, where pins 4 and 5 go to one side of the heater and pin 9 to the other?  The EL-84's should be wired the same with 4 on one side and 5 on the other.  The voltage should be 6.3VAC from one side to the other of the heater winding, or 3.15 VAC from either side to ground.  Try a 100 Ohm 1W resistor from each side to ground.

The battery is not getting the tubes to full performance if the filament is not lit.  You can get distortion and amplification in this condition but the cathodes won't emit the volume of electrons if it doesn't get to design temp.  It will almost certainly be noisier when fully heated.  IMHO I don't believe your results will be the same when you hook up a full performance DC supply.

[ThermionicEngine]

All of these amps have this problem of hum and buzz.The hiwatt fix was 2 15k cathode resistors instead of 1.5k.It does not work.

It may simply be a  bad layout or a botched power transformer run maybe.Grid wires are shielded.
It's not something like that.It doesn't appear to be coming in on the ground of the pcb.The tracks were cut and though helpful didn't do much same with twisted heaters etc.Doubtful if it's a filter cap problem, it was checked out by some repair people who didn't know what to do other than bias it at -25 I think it was.

I f the tubes are not delivering the gain they should would the distortion go down?In this case everything was as it should be. It was still a high gain amp with guitar plugged in.I'm not sure what pin 4 on the el84 is doing but I took a shot at drawing the heater connections.
Only thing that's really worked is the battery.
Any tests I could do on that center tap?

Quick test:
I measured the resistance (with amp off )
from the disconnected heater wires to the earth bolt where the center tap supposedly is.
One yellow wire gives me 0 ohms.
Other one gives me 37k.(I guess this one is going through the diode?)
[/quote]

Hi Frank  We need to stop here for a minute.  Can you upload a photo of the wiring in the chassis?.  Transformer wiring schemes vary but there is some convention out there.  Usually two yellow wires are the heater supply for a 5V rectifier tube.  Green wires usually denote a 6.3 V supply, red wires are the HT and a red striped wire is the center tap fpr the HT winding.  A green wire with stripe often denotes a center tap for the 6.3 V winding.  There is sometimes another tap for the bias supply.   Of course there are plenty of exceptions.  when I hear there are yellow wires and green wires, I suspect the scheme I described is in use.

Unplug the amp for the following.  If yellow is the 6.3 wiring, disconect the wires and measure resistance across them.  The resistance should be very low- .5 to 1.2 Ohm would not be out of line.  Both wires should show no continuity to ground if there is no center tap for the 6.3V supply.  If there is CT for the 6.3 V supply, both wires whould show continuity to ground (if connected or to the wire if disconnected.)  If you can safely do it, try to get AC voltage measured from each secondary wire to ground with the amp on.  If you re not experienced in doing that, please don't OJT now.  A picture would be worth K words at this time.

Putting large resistors into the cathode does the same thing as a current limiter and should bring the gain down.  I do not see this method being a good one.  Typical cathode resistor values are between 1.5K and 2.7K Ohms.  sometimes bypassed with a polarized capacitor with values ranging from .47 Microfarad to 25 Mf.  If we can see those measurements and a photo, one of us might be able to get  handle on what is going on.  If that doesn't work, we'll call in an old Priest and a young Priest and try an exorcism.
Regards,

Larry

[frank57]

Changing the cathodes to 15k on v2 doesn't solve the problem.

You can see the heater setup on the back of the tube pcb.
Here are some shots of the wiring .
Both 6.3v wires beep continuity to the center tap wire which I unhooked if I take the measurement before going through the diode on the power board.
The ohms is 1.2.
Only one wire beeps at the end .It's really two wires the ones from the transformer and then the ones from the turret.
There I get 37k ohms.
The other does not but it's going through a diode to reduce the voltage.

The voltages hooked up were 3.85 and 7.3 without the diode to reduce it.
Now it's 3.85 and around 6.7.
A power board was added and the standby switch was improved.
So if it isn't a center tap it is connected to those heater wires.
In which case what is it?
Sounds like it is a center tap.

Otherwise find a battery than can supply:
So it would be 300 x4 =1200 plus the jj el84 I believe it's 760ma x 2 =1520.
So a total of 2720 milliamps or 2.72 amps?

Typical 6v lantern battery is 11000mAh

[kagliostro]

Some question for you

Have you try to remove the green twisted wires using a temporary short bridge (of twisted wire) to connect from the diodes to the tube ? (to stay away from AC mains input)

The wires (yellow wires ?) outcoming from the transformer and going to the diodes are strictly twisted ?
Try to twist the yellow wires, connect directly one of the yellow wires to one of the green wires, the other you can connect to the other green wire with the two diodes in the middle (cover junction and the diodes with some heat shrinkable sheath, don't use the turret connection and twist strictly the yellow and green wires

You must also consider that on the PCB were tubes are placed you can't use the twisted wires trick,so, if there is a bad track heater layout on that PCB, with AC the solution is only you cut all the heater tracks on the PCB and connect the tubes via twisted wires, that way is better you give DC

Try this:

Find someone that has a DC laboratory supply, disconnect the green wires at the diodes side, set the laboratory supply to the proper voltage and connect to the green wires, what will happen will give you a definitive answer on the use of a DC power supply for heaters

Kagliostro

[frank57]

Yes the yellow wires going to the turret board are twisted as best as possible.
Twisting the wires won't fix it though.
I had to make them longer to reach the board.
They were originally untwisted.It did not do much.
Originally that power board was not there and the heaters went directly to the tube.
I don't think going straight to the tubes with the diodes will do anything.
I could ask a tech nearby if he could do the dc test and if he has a a DC laboratory supply

I did try turning the power transformer at one point and it had some effect on the buzz but the hum went up.

[kagliostro]

Yes the yellow wires going to the turret board are twisted as best as possible.

????

May be they are strictly twisted on the amp, not in the photo

Twisting the wires won't fix it though.

Remember you can't twist the PCB Tracks

Kagliostro

[kagliostro]

I've find the attached image, don't know if is exactly your amp

please summarize what are the wires (1 x black - 1 x green) coming out of the Power Transformer

the others I suppose

Yellow = AC heaters

Red = B+

Kagliostro

p.s.: Try also to twist and put away from circuit all the mains AC wires that are inside the chassis

[frank57]

Black wire goes to the negative of the 100uf on the sideboard.
green wire is center tap for heaters.
The only ac wire that's still near v1 is the one from the output transformer.
But I moved it further away and not much happened.Maybe hook it up somewhere else.
Some hum was there from the power transformer red wires in the original layout but it wasn't the problem.
Twisting the wires won't do much.
In the original amp I twisted the wires and it did nothing.
 It could simply be a very bad pcb layout too wrecking everything or something is wrong on the cathode of v2b.
But most pcb high gain amps tend to have dc power or elevated heaters I think.But it could be a dead end.
Two things for sure:
go around v2b and most of the problem is gone.
Battery test worked but is it good.
Any other way to tell if the heaters are the problem?
Some heater stuff I found.
I think there are some errors in the values here and there.

[frank57]

Could this power the amp?It has 3amps.
Cut the plug and if there are two wires there hook it up to the tube pcb?
http://www.ebay.com/itm/NEW-6V-DC-3000mA-3A-AC-Adapter-Power-Supply-5-5mmx2-5mm-/270793485567?pt=Other_MP3_Player_Accessories&hash=item3f0c8cacff

[kagliostro]

If the green wire is connected to ground try to disconnect it

If you have a 3.15v-0-3.15v transformer (may be your isn't because usually the diodes are used to lower the heater voltage if connected that way) with the CT connected to ground and diodes only in one branch of the circuit you in some way umbalance it

Try what happen if you disconnect the CT (green wire that comes out of the PT - assuming you have a yellow-green-yellow arrangement)

if you haven't good results try what happen if you put (with the CT connected to ground) also in the other branch the diodes to lower the heater voltage

also try disconnecting the CT and put the diodes in each branch (pay attention to put the two of each branch on the right sense)

You can try to add a Humdinger (a pot wired as explained by Merlin, with the wiper connected to ground)
read here http://www.freewebs.com/valvewizard/heater.html as to understand what I mean (remember to disconnect the heater CT)

Could this power the amp?It has 3amps.

very often those little supply aren't so good, probably there is a switching circuit and can't be sure about it

as to be sure about a DC supply try with a laboratory supply

Only one wire beeps at the end .It's really two wires the ones from the transformer and then the ones from the turret.
---------------------------
Now it's 3.85 and around 6.7.

please try to explain one other time (english isn't my language ...........)

Kagliostro

[frank57]

Maybe a board like this can work?
http://www.pmillett.com/DC_filament_supply.htm

I'll try to unhook the center tap today and see what happens.
The voltages are actually a little worse now.
6.9 pin 9 to 4/5
3.9 pin 4 or 5.
Clearly the power transformer is just not doing its job.

[sluckey]

The voltages are actually a little worse now.
6.9 pin 9 to 4/5
3.9 pin 4 or 5.
Clearly the power transformer is just not doing its job.

Look good to me.

Completely disconnect your pt filament leads (including center tap) and reconnect your battery. Now measure the voltage between pins 4 and 5 of each power tubes and between pins 4/5 and 9 of each little tube. What are the numbers?

[frank57]

I did leave the ct hooked up to the the bolt when I did the battery test before, was that an error?
I wasn't getting any reading I think on those 4 and 5 pins or very low.
I tried something just now:
With the filaments back in but the green ct unhooked the hum and buzz doubled.
When I touched my multimeter red to the free end and the black lead on the bolt the bzz hum went down some
but was not eliminated that's for sure.
I also get a dc reading on the green wire of 1.42.Does it mean anything?

[ThermionicEngine]

Greetings all,  The pics help.  It appears to me that the rectifier board not only provides B+ but it also provides a DC heater supply for all 4 12AX7 tubes!  Unless my eyes deceive me (and they do from time to time) the yellow wire goes to a high current rectifier diode.  Then the green wire takes the heater supply to pins 9 and 4-5 of V-4 which alegedly are connected in parallel to the heaters of the remaining pre-amp tubes.  The power tubes get the AC heater supply from the yellow wires.  As I mentioned earlier, the power tubes should humbuck any hum from the heater.

If my observation is correct, and the diode is functional, you should be able to measure DC voltage on the heaters.  It appears in the photo that only one high current diode is in use.  I don't think that would be an effective rectifier.  I would normally see a bridge rectifier in this circuit.  Or at least some type of full-wave rectifier. With the center tap grounded, one yellow wire would go to the anode of each of the big diodes.  The two cathodes (marked with a white stripe) would be tied together and connected to one green wire.  The other green wire is connected to the heater and ground.  If this circuit is properly connected, the heater is probably not the source of hum.

[kagliostro]

To me the connection of heater looks like the attached image plus CT connected

May be I've misunderstand what I've seen

Kagliostro

[frank57]

Well I wish there was a dc power on that board but maybe it could be done somehow.
Kagliostro is right.

There is one filament going through the diodes.
There is an imbalance on the heaters being caused by the diodes on the separate power board.
Although the voltage from 9 to 4/5 goes down to 6.76 from 7.3
it's causing problems elsewhere.
Which might explain why that whole powerboard was not helping anything.
Going into the diodes the filaments are 3.89 dead even.
Coming out one is 3.85 the other 2.9.
This is what you get on the power tubes to ground:
v5
pin4  2.90
pin 5 3.85

v6
pin4 3.85
pin5 2.90

But I have to stress this, the hum buzz was there before as well.
Also I drew out the heater connections better.

[frank57]

Here's the odd heater scheme

[frank57]

Would elevating the heaters like this work or kaboom?
Not sure where to connect the board.
At first filter cap?

[kagliostro]

Before all try to balance the heater branch

try to put two diodes also in the other branch

this will cause the voltage to be equal in the two branch and only 0.7v less in total

Kagliostro

[frank57]

It's a bit of a problem because I don't have any turrets.
Could a resistor be used?
Otherwise I could get rid of the board altogether I suppose.

[ThermionicEngine]

To me the connection of heater looks like the attached image plus CT connected

May be I've misunderstand what I've seen

Kagliostro

Hard to see.  The way I'm guessing or your view is still a half wave rectifier.  I don't think that was the intent of the design.  Maybe it has been changed.  If each yellow wire is conected to the anode of the big diodes and the cathodes are tied together to one green wire you have a DC supply.  A cap would be indicated as well, Maybe start with an E-lytic 30 mfd @ 50 VDC Negative to ground. Cheers

[kagliostro]

Hard to see.

here you can see better

http://www.el34world.com/Forum/index.php?action=dlattach;topic=12208.0;attach=25451

only one of the wires is connected to the diodes and the two diodes are connected paralleled and reversed each other, that way there is only the reduction of the voltage about 0.7v, and you have AC not DC

Kagliostro

[frank57]

What about an idea like this?
Use the turrets and make a bridge rectifier.Although I'm not sure which diodes to use.
Maybe put some resistors somewhere there to bring the voltage down to 6.3?
I think it will go up to 8vdc.

For the elevate the heaters where do I hookup: to the first filter cap +?

[ThermionicEngine]

That should get you a DC heater in the pre-amp.  You can use 1N5408 diodes if they will fit on your board- they are 3 amp diodes and should handle the current easily-  I'm guessing 1N4007's will do but I would use 5408's                                   

[frank57]

Is heat an issue and are resistors needed to get the voltage to 6.3?
I could get a diode bridge and hook it up outside the amp to the filaments and so on and see if it works first too.
Would the power tubes be ac or dc?

[frank57]

I tried an artificial ct after the diodes on 4 and 5 of the last el-84.
This has brought the hum down quite a bit.Nothing for the buzz though.
The voltages are now 3.4 on 4 and 5 and 9 to ground and 6.8 on 9 to 4/5.
So does it mean the original ct is a piece of crap?

It was balanced as far as the voltages at 3.8 everywhere and 7.7 on 9 to 4/5.
Or would simply adding another pair of diodes have done the trick?

I'm wondering here if a big part of the problem on these amps is simply a crap
buzzy hummy Power transformer on top of whatever else is off.