Benson Monarch

[Aaron]

Sorry, I hadn’t been following the thread. I will get the amp back and get more pictures.
Thanks,
Aaron

[chopstuck]

I tried 33 mfd at all 3 nodes with minor improvement. Increasing the 1k resistor value between the nodes and moving the pi down to the third node helped the noise. Just like a tweed amp. Changes the response though..

[pullshocks]

Sorry, I hadn’t been following the thread. I will get the amp back and get more pictures.
Thanks,
Aaron

That would be great. Could you confirm whether that the input and output jacks are grounded to the chassis, i.e. no isolating washer.

Could you measure the phase inverter balance pot resistance setting.

Also could you listen for hum and tell us what you hear.

Thank you

[dorrisant]

Here are the possible F&T sizes for comparison...

F&T Cap Sizes.png

...and this shows 100µF-450v and 220µF-300v caps next to a couple of types of octal sockets as well as a 40/20/20/20 JJ can cap. I've seen it in print somewhere that JJ caps were used. I'm assuming this was said because it would be visible from the outside of the chassis. Has anyone seen a 20/20 or 20/20/20 from JJ?

Filter Possibilities.jpg

Looks like there are several possibilities.

In this pic it looks like a two section can-cap with a dropper between them. The two axial caps look they could be paralleled, grounds are tied anyway. Other pics show one cap here, that's why I make that assumption. Datasheets for the 5AR4 state a maximum of 60µF for the reservoir. I know this is abused but how far could this be stretched to get the most filtering?

Offhand, I think the Rocket uses 80µ without any trouble.

Benson Monarch Filters.jpg

[T Wilcox]

Good point about max capacitance on the reservoir for tube rectifier! and your pic of the F&T sizes leaves me thinking more like 47uf than 100uF as I stated earlier but it will be great to find out for sure
@Aaron Thank you for jumping back in!

[didit]

In this pic it looks like a two section can-cap with a dropper between them.

A PI filter in other words.

Other pics show one cap here, that's why I make that assumption.

The other pictures appear to have a "can" cap, which would also provide sufficient capacitor elements to support a front-end PI filter.

(edit)

Explanatory reinforcement: a pair of 30uF caps with roughly 100 Ohm resistor betwixt provides better AC ripple filtering than simply a single 60uF and has less rectifier stress, with a small drop in overall supply voltage.

Best .. Ian

[mojotom]

From pictures the Monarch uses a dual can cap so two filter capacitors plus another node with a single cap (or two in //).
No PI node then (a 4th cap would be needed).

I would say a dual 47uF for B+ after the standby and screens then another node for the rest of the amp, either a single 47uF cap or two 22uF/500V on // (about the same capacity).
On one picture I could see a 1k and a 10k power resistor.

Could be wrong, just a guess.

If the point is to lower the hum you could solder some caps in // to you existing cap until hum become acceptable.

Everybody building one seems to love the sound of it.

Do you find the US/U.K. switch useful ?
Could be removed or a nice add-on ?

If you take a look at the Benson preamp stompbox schematic there’s an alternative Bass control (and bright cap) that might be worth trying (I think it might be similar to the Chimera and maybe Nathan tonestack).

[dorrisant]

In this pic it looks like a two section can-cap with a dropper between them.

A PI filter in other words.

Other pics show one cap here, that's why I make that assumption.

The other pictures appear to have a "can" cap, which would also provide sufficient capacitor elements to support a front-end PI filter.

I just want to be clear... All of the pics that I have seen have a can-cap in roughly the same location. Some have a single axial cap, some have two. That is factual.

My opinion is that since some pics have a single axial and some have two, these could be two filter nodes or in parallel for one node. I was thinking that this axial setup could be the reservoir. It could be two half-values in parallel or just as a single if the cap meets the voltage spec. I was thinking that the two axials would be used for the reservoir and an additional filter node for the reverb builds and a single axial for non-reverb builds but this theory won't hold water. This pic is obviously a head, determined by parts placement. It is also a non-reverb head due to the fact that none of the reverb circuit parts are there. Yet it has two axials and at least two nodes on the can-cap. So it could be three or four nodes, depending on how the positive ends of the axials are connected.

Benson Monarch under 4.jpg

I'm not trying to argue with anyone, I'm just using logic to narrow down the convention. For comparison, this pic that Aaron posted has the reverb parts and is obviously a combo layout. We'll have to wait for more pics to determine how many axials were used, unless Aaron posts confirmation of either way.

Benson Gutshot.gif

[dorrisant]

From pictures the Monarch uses a dual can cap so two filter capacitors plus another node with a single cap (or two in //).
No PI node then (a 4th cap would be needed).

I would say a dual 47uF for B+ after the standby and screens then another node for the rest of the amp, either a single 47uF cap or two 22uF/500V on // (about the same capacity).
On one picture I could see a 1k and a 10k power resistor.

The two 47µF caps in parallel would be roughly 100µF... I don't know if I'd run a 5AR4/GZ34 that far over spec. Maybe as a test, but in a production amp... ? It is not like the new production 5AR4s are so reliable that it wouldn't matter... right? It could be that the can-cap has the reservoir and 2nd node with a large wattage dropper between them and the axials are for the preamp. Datasheet says that 60µ is max for the reservoir, but that doesn't mean that the later preamp filters have to be 60µF or less. They can be larger than the reservoir, you just don't see it that often.

Again with this pic, but the dropper across the can-cap looks like a 1k. Putting this pic with Aaron's schematic, can we reasonably assume that the can-cap contains the reservoir and 2nd filter stage?

Benson Monarch Filters.jpg

Still not trying to stir shit, just trying to be logical.

I still can't find a JJ can-cap that is a dual 20-22µ. Closest is a 40/20/20/20 but it doesn't look right for the pics. You can find 16/16 and 32/32, but I can't seem to find anything close to 20/20. I could be chasing an erroneous point. Can anyone confirm this? Better: does anyone have a picture that shows the topside of the chassis where the can-cap is visible... like 'read the label somewhat' kinda visible?

As a side note: this sure reminds me of the time that the Xits X10 was up for debate. Who remembers that?

[dorrisant]

I will update my layout when a few more details are clear.

A big thank you to chasbenson... Boy how I love a good puzzle, and you sir have created a beauty!! Better than a Hanayama level 6.

[mojotom]

I meant the dual cap could be a 47uF used for B+ (one section, appropriate for a new production 5AR4 or almost on a NoS 5Y3) and the other section for the screens node after the 1K “choke” resistor.
Most dual can caps values are 33u, 47u and 100u, not much in between unless a few here and there. Could be custom ordered too.

Then another 47uF cap for the rest of the amp (that F&T single 47u or 2x22uF, both at 500V look the part (still just a guess on my part for the fun of it). Could be 100uF (or 2x47u) too.

[Aaron]

Sorry folks,

I looked at the schematic again and realised I had kept the original heater wiring, I traced it first before repairing it. The reason I had the amp is that the pre amp heater supplied had failed. The attached schematic shows what I did to it.
The filter caps were 2 2x20uf JJ Cap Cans, the second half of one wasn't connected.
I'm getting the amp back next week.

Thanks,
Aaron

[didit]

A picture (annotated) perhaps better than some more words ..

PS.jpg

[mikeywoll]

I tried 33 mfd at all 3 nodes with minor improvement. Increasing the 1k resistor value between the nodes and moving the pi down to the third node helped the noise. Just like a tweed amp. Changes the response though..

Your post caught my eye...

Basically every interview with Chris Benson mentions he was inspired by old Supro or Ampeg designs. These old amps sent the first node to the OT, used a 1K voltage dropping resister between the first and second node, ran the screens off the second node, then used a 10K to 22K resister between the second and 3rd nodes. Every pre amp tube, including the phase inverter, ran off the 3rd node. Did you try this configuration, without increasing that 1k resistor?

I'm a few days behind, and building with a SS rectifier so I expect to make some changes vs tube rectifier, but my plan was to try the configuration above. If it buzzes, I plan to add an additional node and resistor so I can shift the V1 and Reverb tubes to an isolated node, leaving the PI alone on the 3rd node.

FWIW it seems really suspicious that Benson would choose to put 4 electrolytic caps in the amp but only use 3 (the 2x can and then either 2 caps or a 2in1 depending on the picture). I wonder if we are missing a 4th node on the schematic.

Thanks to everyone contributing to this thread. It has been super interesting to follow along.

Mike

[chopstuck]

Thank you about the tip for grounding the DC heater Aaron. I have discovered that the 0.22 vac leaking from the PS into the second node is getting into my PI output.
I tried adding another node previous to my string but that just dropped my voltages too much. I added another 20 mfd cap to the second node feeding the PI and decreased the ac ripple from .26 to .012 vac without changing much else. This dropped the vac going to the grids of the power tubes by 2/3s at idle.
From .051 to .022 vac ! This resulted in much less hum but it's not good enough. It hums more than a cheap SE amp.

I'm sure my noise is PI centric so , I will have to modify the PI for better balance. IV tried adding more resistance to the fixed side but that hasn't helped.