Bogen CHB100 / Ampeg B15... maybe.

a fine line between stupid and clever

Bogen CHB100 / Ampeg B15... maybe.

Postby ioginy » Tue May 03, 2011 9:34 pm

Good day all. I am doing a little project that I wanted to get some eyes that aren't mine to take a look at.

The voltages are ideal estimates, so I'll worry about that after it starts getting built. The transformer adds up, so it should be dandy.

Image

Thanks in advance, you guys are always such a great help, and wealth of knowledge.

Cory.
User avatar
ioginy
KT88
 
Posts: 272
Joined: Sun Oct 18, 2009 5:38 pm
Location: Edmonton, ab

Postby soundmasterg » Wed May 04, 2011 4:01 am

I didn't look super close to see if you have it, but the stock CHB100 uses a voltage doubler in the power supply for the B+ and is about 480v with today's wall voltages. I've been modifying one myself and that is where it idles at. Mine makes 58 watts RMS and 98 watts peak with four 7868's and the new circuitry I've added around it.

Greg
soundmasterg
 
Posts: 76
Joined: Thu Feb 08, 2007 5:35 pm
Location: Oregon

Postby dcgillespie » Wed May 04, 2011 6:16 am

Three things you might consider:

1. Add a high value resistor from the grid(s) of the 12AX7 power amp input tube to ground -- in case the effects loop jacks lift and/or an external source does not provide a DC return path to ground for the grid. A resistor of about 1 Meg would be good.

2. The plate load resistor for the same stage is unusually large for the amount of bias you are providing -- particularly with the two halves paralleled. If the overall resistor is providing the DC operating point you want, it would likely be much better to add a decoupling cap to the circuit. That way, you could use a 390K dropping resistor from the 370 volt source, decouple, and then use a 47K resistor for the actual plate load. The overall DC conditions would remain basically unchanged, but the AC conditions would make more sense. However, even with this change, with the bias level you have chosen for this stage, I think the dropping resistor will end up likely being no more than about 180-200K to achieve a reasonable DC operating point.

3. Using a 1K resistor in series with the plate lead of the top and bottom output tubes is way-way over kill, causes a definite loss of power, and endangers the screens of those two particular tubes. To control parasitic oscillations, a value of 10 ohms is completely adequate at these locations, and much more appropriate for the application.

Good luck with your project!

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby ioginy » Wed May 04, 2011 9:51 am

The power supply is essentially exactly the same as the Bogen. I figured to get the posted voltages, it would be best not to mess with something that is tried and true for this design.

I added the 1M grid resistor, excellent point on that one.
The 10 ohm resistors to the 7868 screens has been changed as well, but I was wondering if I could get a little more info on that. The original bogen schematic uses 1K, which is almost exactly how the output is designed. Also, any amp I have ever built or looked at has always had 1K or 1.5K. It seems like an odd adjustment to make.

http://makearadio.com/schematics/images ... 100b-6.jpg

2. The plate load resistor for the same stage is unusually large for the amount of bias you are providing -- particularly with the two halves paralleled. If the overall resistor is providing the DC operating point you want, it would likely be much better to add a decoupling cap to the circuit. That way, you could use a 390K dropping resistor from the 370 volt source, decouple, and then use a 47K resistor for the actual plate load. The overall DC conditions would remain basically unchanged, but the AC conditions would make more sense. However, even with this change, with the bias level you have chosen for this stage, I think the dropping resistor will end up likely being no more than about 180-200K to achieve a reasonable DC operating point.

I didn't really have a set bias point for this 12ax7. The only real reason I am putting them in parallel is because I felt bad about leaving an unused half in there. From what I could find, when paralleling 12a?7's the plate resistor should be double what it was for a single triode, and the cathode should be half.
If there is a more appropriate layout for this tube I would love to change it up. It always sat just a little wrong in my head, but I couldn't figure out why.

Thanks.
Cory

updated schem.
Image
User avatar
ioginy
KT88
 
Posts: 272
Joined: Sun Oct 18, 2009 5:38 pm
Location: Edmonton, ab

Postby TomMcNally » Wed May 04, 2011 11:03 am

Sorry to butt in, Cory, the Bogen schematic you posted shows
15 ohm, 1 watt where you have 1K ??? (on the plates)

The 1K on the grids are fine as is.

... tom
User avatar
TomMcNally
Darling du Jour
 
Posts: 2729
Joined: Sat Nov 19, 2005 2:19 pm
Location: Northfield, NJ

Postby ioginy » Wed May 04, 2011 11:50 am

*smacks forehead*

And I feel like a schmuck! Thanks for pointing that out, I should know better than to start ifddling with schematics when there is still sleep in my eyes. :))
User avatar
ioginy
KT88
 
Posts: 272
Joined: Sun Oct 18, 2009 5:38 pm
Location: Edmonton, ab

Postby dcgillespie » Wed May 04, 2011 1:26 pm

Your first schematic shows 1K resistors in series with the plate leads of the top and bottom output tube, with no resistors provided for the screen grids at all. Your second schematic still has the 1K plate resistors in place, but now has changed the 1K control grid stopper resistors to 10 ohms.

The grid stopper resistors should stay at 1K as they originally were. It is the plate resistors I questioned. The Bogen schematic you provided show the use of 15 ohm resistors in this position, where I was suggesting 10 ohms. Either would be fine.

While not shown on either of your schematics or the Bogen schematic, it would be a very good idea to add 100 ohm screen stability resistors in series with the screen grid terminal of each output tube. These can be 1/2 watt resistors, and will minimize any tendency towards output tube arcing, which the original Bogen design could promote under heavy transient conditions.

When you parallel two triode sections, it is customary to halve the plate load resistor, cathode bias resistor, and even the grid resistor. Noise will go down, and current flow will double. To minimize the possibility of parasitic oscillations, it would be best to use two separate grid input resistors (one for each grid) between the grid and the new 1 Meg grid return resistor. A value of about 2.2K in series with each grid should be sufficient. Also, with the current flow of the stage doubled, the dropping resistor in the power supply will likely require adjustment to maintain the original supply voltage value.

Finally, since this stage operates within the FB loop, you may need to adjust the 15 PF phase advance cap in the feedback network to maintain proper stability. Also, the original Bogen schematic had a 220 PF and 82 PF effectively shunting HF portions of the signal to ground from the output of the first AF amplifier stage, before proceeding to the phase splitter. Components like these are very important to the stability of a FB amplifier, so you should first include them before determining that you don't in fact need them. By paralleling the input triodes as you have, the frequency response of the stage has been extended, so that all of these components may require adjustment.

Your basic approach is a good one, but the design changes you are making will likely require a scope and square wave generator to obtain the best results.

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby ioginy » Wed May 04, 2011 10:24 pm

Thank you for the detail on that one. All the changes suggested are exactly what I am looking for.

Here is an updated schem. The second dropping resistor will need to be a bit of an experiment when I get it built to find the proper voltage (unless someone knows of a calculator).

Image

I sadly don't have a scope, or a sine wave generator. Is there a way to do it mathematically. I'm no math genius but I'm willing to put in the time if it will improve this project.

Oh, I should also note that this is intended to be a bass amp, so i tried to make all my adjustments with the goal of low distortion. The bias switch on the first stage is to increase gain to suit music style. The original value is the middle, the first being a standard fender bias and the the third (i feel) the maximum value for low gain but good level.

-edit-
I also just realized that the filter caps are posted as the original 10uf. I would like to increase them to 40uf ideally, but have 20uf's as well if someone thinks it might turn things bad, though I can't think of any way in which it would.
User avatar
ioginy
KT88
 
Posts: 272
Joined: Sun Oct 18, 2009 5:38 pm
Location: Edmonton, ab

Postby dcgillespie » Thu May 05, 2011 8:14 am

Your circuit is starting to clean up nicely now. A couple more points to consider:

1. Regarding your B+ dropping resistors, the first resistor supplying the phase inverter and AF amp stage will probably not need any change at all. The second one supplying the preamp will likely need to be increased to hit your target voltage. I would start with 15-18K and go from there.

2. With the 12AX7 AF amplifier stage paralleled for the power amp, the cathode bias resistor for this stage has been basically cut in half. This maintains the original operating point for the stage, but as a last detail, also cuts the NFB level in half as well. To restore the feedback to its original level, the FB resistor should be reduced to 22K, and the phase advance cap placed in parallel with it should be increased to about 32 pf. This network, in tandem with the new 220 ohm cathode resistor, will then return the FB level to its original design level. The other HF response caps (220 and 82 pf) can remain the same value as they were.

3. The output transformer will determine the success of this amplifier as a bass guitar amplifier. Typically, amplifiers intended for public address service have little low frequency response, since little low frequency response is needed. As a result, the output transformer can be made rather economically, since it does not have to pass much significant low frequency content. In this case, economical equates to a smaller core, and therefore, smaller size. Since this amplifier will produce about 60 watts RMS output, it might be good to compare the physical size of the transformer in this amplifier, to that of a known 60 watt tube bass amplifier.

Remember the original Sunn bass amplifier that used a direct knock-off circuit of the Dynaco MK III power amplifier? That amplifier literally used a Dynaco A-431 transformer in the original production units. This transformer is likely much larger than the one in your Bogen. Granted, it probably does not need to be as big as the Dynaco transformer was, but today's bass guitars have much lower responses, going well below 30 Hz IIRC. To pass frequencies in that range at full power requires an OPT capable of that performance.

I'm just mentioning this to be aware of the concern. It's not necessarily a deal breaker, but it may imply that the output transformer will require a change for the service you are now looking for the amplifier to provide, since it is not in keeping with that of the original design.

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby dhuebert » Thu May 05, 2011 2:26 pm

I don't understand the paralled 12AX7 circuit at all. Why is it there? What I would do is put the tone control recovery circuit in front of the effects loop and then a cathode follower to the effects loop to minimize output impedance and then a single buffer stage before the PI. Something else to consider is a cathode follower in front of the tone controls, I don't personnaly do this but you see it in many designs. The load impedance of the tone controls change as you adjust them which can have a large effect on the frequency response of the previous stage. For HiFi amps this is an issue but not usually for guitar amps, it is just something to be aware of.

I've had a lot of success with Ultra-linear output transformers and no global negative feedback in my bass amp designs.

The Hammond transformers have excellent frequency response for bass.

Don
User avatar
dhuebert
KT88
 
Posts: 820
Joined: Thu May 01, 2003 9:26 am
Location: Winnipeg Manitoba Canada

Postby dhuebert » Thu May 05, 2011 2:29 pm

Yes, I have to say: running your effects loop straight out of the tone stack is not a good plan. Isolate the tone stack as much as possible. ie impedance as low as possible leading into it and as high impedance as possible loading it. And the same thing with the effects loop.

Pardon me if I sound harsh but this whole design seems heavily compromised; from the power supply circuit to the tone stack circuitry to the output section. I will be very surprised if it is any good at all as a bass amp. It might make a very good guitar amp, tho.

Don
User avatar
dhuebert
KT88
 
Posts: 820
Joined: Thu May 01, 2003 9:26 am
Location: Winnipeg Manitoba Canada

Postby ioginy » Fri May 06, 2011 7:20 pm

*nods* i see what you're saying there. I used the "line out" placement from the b15 preamp, and was hoping to keep the layout simple by just turning it into an effects loop. That being said, I would like it to work well.

I would also like to not have to add any more tubes to the mix. There is the other half of the 12ax7 that I threw in parallel, but can be used for anything really.

Moving the effects loop to the other side of the PI driver is easy enough, but would you have any suggestions in how to impliment your ideas using only one additional triode?
User avatar
ioginy
KT88
 
Posts: 272
Joined: Sun Oct 18, 2009 5:38 pm
Location: Edmonton, ab

Postby soundmasterg » Sat May 07, 2011 3:22 am

The stock Bogen CHB100 output transformer has a 5000 ohm primary impedance with the four 7868's and is about 2/3 the size of the Dynaco A431 output transformer that came in the Sunns. The one I modified is now a guitar amp, but it has plenty bass for me, though for a bass amp I think it might be somewhat lacking. The voltage doubler in the power supply works well for me for guitar purposes but may sag a bit more than you would like for bass. I've increased the caps in the first filtering stage to 220uf, so its the same as a 110uf cap in a normal supply, but you may want to go a little higher for bass. Not too high though as the transformer may not be happy having to supply so much current quickly...it is rather undersized also.

Greg

P.S. >> @dcgillespie, How is a square wave generator and a scope helpful to root out oscillations or optimize changes to an amp?
soundmasterg
 
Posts: 76
Joined: Thu Feb 08, 2007 5:35 pm
Location: Oregon

Postby ioginy » Sun May 08, 2011 10:07 am

I just did some looking, and I think I'll just take the effects loop out. It really isn't necessary for my purposes, and the great thing about building amps is that if it is needed int he future, I can put it in later.

Construction starts today :))
User avatar
ioginy
KT88
 
Posts: 272
Joined: Sun Oct 18, 2009 5:38 pm
Location: Edmonton, ab

Postby ioginy » Sun May 08, 2011 10:45 am

I decided to use the other half of the 12ax7 and put it to use in somewhat of the way dhuebert suggested, but putting a cathode follower before the tone stack, like the old bassman's used to do.

Image \

-edit-
Here is the updated schem, I also changed the power supply because I noticed my caps have much higher voltage ratings than those originally used.
Image
-edit-

The 1M and 2.2K resistors I added to help with the effects loop have been left in, but I have a feeling they will not be needed, at least the 2.2K won't be needed. The 1M to ground seems like a smart thing to have anyway.

Any thoughts?
User avatar
ioginy
KT88
 
Posts: 272
Joined: Sun Oct 18, 2009 5:38 pm
Location: Edmonton, ab

Next

Return to guitar amps

Who is online

Users browsing this forum: No registered users and 8 guests