London Power ad

[-]
Search the Forum








(Advanced Search)

Bogen CHB-20A Rebuild
#21
Hi Guys

I do not believe you will find the "caterpillar" insulator with adhesive; you just have to use the right one for the materials it wraps around. The manufacturer of the grommet should have info regarding what the usual or best glue might be.

Plans to re-do things later almost never actualise. Once something is assembled and wired and functions that is the end of its story. It is just human nature and the world we live in. Time to move on to the next project.

It is always nice to find a local place to order hardware from. Larger cities usually have enough trade to support such places, usually catering to the auto industry and machine shops. In London, there is a place called Facca Fasteners that either had in stock or could acquire pretty much anything I needed. Here in Thunder Bay, one third the size of London, Intercity Industrial Supply can order anything Spanaur offers (they are a hardware manufacturer and distributor). Other places here with "fasteners" in their name or their product range tend not to have such broad sources as they cater to a niche of businesses.

Sometimes you do end up using online suppliers, and as usual, some places are only good for certain things. It is a learning process that is greatly influenced by what is convenient to you or for your locale.
Reply
#22
It does seem to be hard to find adhesive versions of "caterpillar" grommet (this stuff seems to have oodles of names). I did find one, Panduit GEE62F-A-C, so that's the one I'm considering. That's actually what I used in the pictures, but my spool's adhesive expired in 2002 so it's not ideally sticky anymore. I called Panduit and they said that the adhesive they use is a super-glue pressure-sensitive hot-melt type so it seems it should be much stronger new than the expired stuff I'm using.

Agreed about plans to re-do things. I'd take some more time to do it exactly right, however I'm on a time crunch (this is in part for a school project I'm supposed to present in early May) so a few concessions are made with the hope that the itch for perfectionism will be enough to motivate fixing them later.

You're right, local sources are great. I'm lucky enough to be nearby a McMaster will-call location, so I've been ordering online and then picking up in person. I'm also spoiled by several electronics supply stores in the area. I plan to move later this year, and I'm already trying to find where I'll source parts from in the new city. I think I'm going to miss all the local stores I have in my current city (Los Angleles/LA County).
Reply
#23
Ok, I'm putting together a shopping list for some electronics parts I need and would like to double-check with you guys the fuse values I've selected. I checked TUT3, and all I saw so far was current calculations. Useful, but I'm still wondering about margin factors.

Mains fuse is 1A, it's what the stock bogen was spec'd for. and even with the additional 12AX7 and the change to EL84 power tubes that only equates to an extra 38mA if I did the math right. The stock fuse is about double actual stock load if my calculations are correct, so I'm thinking there is enough headroom that the fuse doesn't need to change.

Heater draw is 2xEL84 + 3x12AX7 = 3.32A plus a few milliamps for the pilot light LED, so I'm thinking 3.6A minimum, possibly 4A. The heater is not center-tapped.

Stock it's a 20W amp, so peak audio output power would be 40W, add 20% per TUT3 for waste power and we're looking at 48W. Per the schematic the power transformer is designed for 400V plate voltage at 120V line voltage, so using that number I'd expect peak draw of about 120mA. This is where I start feeling funny. 120mA seems like an awful small fuse value, even if I'm rounding to a 200mA standard value fuse. Is that correct? Am I missing margin to keep the fuse from constantly blowing?

For the bias winding I'll have around 90VDC raw output across two parallel 25k pots in series with a 6k8 resistor to ground, so around 5mA assuming no grid current. I'm guessing just use the smallest fuse size I can, which is 200mA I think?

So, how does that all sound? Another option is to only use a mains fuse, like I think the Soma 84 does. Not quite sure why some of the projects are fused to the hilt and others aren't. I'm going to see if I can reverse engineer the Standard's fusing to back out margins.

Thanks for any input!

P.S. Forgot to mention. There are no center taps, and all the fuses I was talking about are assumed to be slo-blo.
Reply
#24
(05-01-2024, 05:15 PM)physics Wrote: Ok, I'm putting together a shopping list for some electronics parts I need and would like to double-check with you guys the fuse values I've selected. I checked TUT3, and all I saw so far was current calculations. Useful, but I'm still wondering about margin factors.

Mains fuse is 1A, it's what the stock bogen was spec'd for. and even with the additional 12AX7 and the change to EL84 power tubes that only equates to an extra 38mA if I did the math right. The stock fuse is about double actual stock load if my calculations are correct, so I'm thinking there is enough headroom that the fuse doesn't need to change.

Heater draw is 2xEL84 + 3x12AX7 = 3.32A plus a few milliamps for the pilot light LED, so I'm thinking 3.6A minimum, possibly 4A. The heater is not center-tapped.

Stock it's a 20W amp, so peak audio output power would be 40W, add 20% per TUT3 for waste power and we're looking at 48W. Per the schematic the power transformer is designed for 400V plate voltage at 120V line voltage, so using that number I'd expect peak draw of about 120mA. This is where I start feeling funny. 120mA seems like an awful small fuse value, even if I'm rounding to a 200mA standard value fuse. Is that correct? Am I missing margin to keep the fuse from constantly blowing?

For the bias winding I'll have around 90VDC raw output across two parallel 25k pots in series with a 6k8 resistor to ground, so around 5mA assuming no grid current. I'm guessing just use the smallest fuse size I can, which is 200mA I think?

So, how does that all sound? Another option is to only use a mains fuse, like I think the Soma 84 does. Not quite sure why some of the projects are fused to the hilt and others aren't. I'm going to see if I can reverse engineer the Standard's fusing to back out margins.

Thanks for any input!

P.S. Forgot to mention. There are no center taps, and all the fuses I was talking about are assumed to be slo-blo.

It just occurred to me that I doubled the currents for each of the 12AX7 heaters... Oops. So subtract 0.9A from the heater fuse to get about 2.6A, probably rounding up to 3A. And the calculation for the additional current required of the mains winding fuse with the switch to EL84's and an extra 12AX7 should be 22mA, not 38mA. Sorry!
Reply
#25
Hmm, after some thought and further reading, fusing the bias supply may not be the best idea as it adds a failure point that could result in damage to the output transformer. Plus it just trades certain power transformer failures for possible output transformer failure. Seems like something more sophisticated is needed, that'll kill plate (and/or screen?) voltage if the bias supply fails. For now I think I'll leave the bias fuse out.
Reply
#26
Hi Guys

Fuse calculations are given in TUT3 and elsewhere in the series, and also here on this forum:
Fusing for Bias Supply? Fusing for the OT?
post-2

In my amps, I fuse every supply so there is proper protection for each winding of the PT, and that includes the bias supply.

You can always look at every component as a potential breaking point which maybe cause a cascade of failures. On the other hand, you have to put your faith in something. So, say you fuse the bias supply and that fuse opens - it could be a nuisance failure or because of a legitimate fault, does not matter. Have you also fused the OT? If not, then the OT may be in danger. If you have fused the OT as well, then that fuse will open second and protect the $$ items.

You could devise all kinds of interlock approaches between the bias supply and the plate supply, but then you are relying on these extra components. Are they more reliable than the core circuit components? They shouldn't be.

The reliability rates for most components is pretty fantastic. I choose components for reliability and circuits for tone, therefore I do not follow a specific prototype circuit to attain a particular tone.

The bias winding in a PT is more likely to fail because it is of such light construction compared to the other higher-current windings than for any other reason. Vibration during normal operation can break the wire and then the bias winding no longer functions. I've had this happen on larger PTs and on subsequent runs spec'd the bias winding to be much heavier itself to have a margin against a repeat failure. Really, it was a fluke of manufacturing that lead to that failure.

The tensioning of the wire during manufacture is called "scaning". I had a small audio transformer fail because of a scaning issue - very tiny wire and many many turns. Talking to the manufacturer, I determined that I could use the rest of the batch by adding one resistor to the circuit which did not effect the basic performance but protected the transformer. Another manufacturing issue but also another transformer!

Since transformers are so expensive, i have no hesitation to add tons of fusing to protect them. You have to abandon the larger fuse sizes and go with 5x20mm or even smaller. A friend of mine uses pico fuses that solder to the PCB. I can't say i like that method but for what he builds he has never had a fuse failure and really nothing he builds is going up in smoke anyway Smile

Have fun
Reply
#27
(05-02-2024, 12:37 AM)K O'Connor Wrote: Hi Guys

Fuse calculations are given in TUT3 and elsewhere in the series, and also here on this forum:
Fusing for Bias Supply? Fusing for the OT?
post-2

In my amps, I fuse every supply so there is proper protection for each winding of the PT, and that includes the bias supply.

You can always look at every component as a potential breaking point which  maybe cause a cascade of failures. On the other hand, you have to put your faith in something. So, say you fuse the bias supply and that fuse opens - it could be a nuisance failure or because of a legitimate fault, does not matter. Have you also fused the OT? If not, then the OT may be in danger. If you have fused the OT as well, then that fuse will open second and protect the $$ items.

You could devise all kinds of interlock approaches between the bias supply and the plate supply, but then you are relying on these extra components. Are they more reliable than the core circuit components? They shouldn't be.

The reliability rates for most components is pretty fantastic. I choose components for reliability and circuits for tone, therefore I do not follow a specific prototype circuit to attain a particular tone.

The bias winding in a PT is more likely to fail because it is of such light construction compared to the other higher-current windings than for any other reason. Vibration during normal operation can break the wire and then the bias winding no longer functions. I've had this happen on larger PTs and on subsequent runs spec'd the bias winding to be much heavier itself to have a margin against a repeat failure. Really, it was a fluke of manufacturing that lead to that failure.

The tensioning of the wire during manufacture is called "scaning". I had a small audio transformer fail because of a scaning issue - very tiny wire and many many turns. Talking to the manufacturer, I determined that I could use the rest of the batch by adding one resistor to the circuit which did not effect the basic performance but protected the transformer. Another manufacturing issue but also another transformer!

Since transformers are so expensive, i have no hesitation to add tons of fusing to protect them. You have to abandon the larger fuse sizes and go with 5x20mm or even smaller. A friend of mine uses pico fuses that solder to the PCB. I can't say i like that method but for what he builds he has never had a fuse failure and really nothing he builds is going up in smoke anyway Smile

Have fun

Thanks for the info Kevin! 

I dug through TUT3 and all I found explicitly addressing fusing values (besides stuff like "an 8A heater fuse lets us use different valves") is the appendix on the universal fusing method for power transformers. Did I get an older revision of TUT3 and is there a newer revision with more fuse details or am I just blind? That other post you mentioned said that TUT3 called out specific fuse holders, but I didn't see that when I went through TUT3 last night so I wonder if mine is missing some stuff. Granted, I was tired and it could have been a call-out on a schematic that I missed. I'll have to dig through the rest of the series to see what stuff I read and forgot about fuses.

That post you mentioned was really helpful, not sure why I didn't find it early. Guess I need to look harder. Using the 1.25*1.25 fudge factor mentioned, my heater fuses are 0.3A too small, so I'll have to swap those out. Bias supply and HT winding fuses are fine by virtue of not being able to get smaller ones than 250mA at the shop, which is 2x my non-fudged HT calculation and oodles more than the bias draw. 250mA is definitely more than what my bias transformer can handle though... Maybe I should find a smaller fuse. Looks like Digikey has some small current value ones that should probably do it.

I see your point on picking something to trust. For some reason I had it in mind that an interlock of some sort would be better since I can program a specific current value compared to my impression of fuses being a bit cruder. Though thinking about it some more, I'd think a fuse is the simpler, less-moving-parts protection method. Ideal might be an interlock in addition to fusing, for redundancy. Also, like you pointed out, the output transformer should be fused anyway since there are other ways to kill it besides the bias failing. Time to get more fuse holders...

And yeah, 5x20 seems like the way to go. I was wrestling with getting some 6x30's to fit on my board since I had some scrap clips for those, but my life got so much easier when I switched to 5x20 holders I just picked up. Way more space. I can't say I envisioned there being quite this much mechanical design and problem solving involved when I started this project... Not that I'm complaining.

Thanks for the help!
Reply
#28
I think I got fuse locations worked out on the PSU board. Left to right is bias, heater, and plate windings' fuses. I've drilled the holes, and test fit the eyelets and fuse holders.
   
Since I'm hanging an auxiliary transformer off the heater winding to use for the bias supply, I'm thinking that the auxiliary transformer should come after the heater fuse so that if the aux transformer fails short for some reason then the heater fuse will trip before the heater winding goes up in smoke too.

Anyhow, now gotta drill out the holes for the other components, stake all the eyelets, and solder!
Reply
#29
Drilled a few more holes for the voltage doubler bridge, and swaged all the eyelets I've drilled for so far.

It took some doing to figure out how to get a good swage without folding the edge of the eyelet in, collapsing it, or anything like that. I'm using Keystone's #44 eyelets and staking tool #1715 from Keystone. Initially just using the tool and a hammer it was near impossible to get a good swage. I attribute this to the fact that the 1715 tool is just barely big enough for the 44 eyelets, so unless I have everything perfectly square when tapping on the hammer the rolling portion of the tool surface contacts the edge of the eyelet before the conical section of the tool can flare out the eyelet so that the rolling surface actually rolls down the eyelet instead of collapsing or folding the wall inwards. I tried using a hand press I picked up cheap, but it seems like there is too much slop in it to keep things perfectly square, so the same issues occurred. I found that using a center punch of the right diameter, I could pre-flare the eyelets enough that when I use the staking tool manually the eyelet walls will cleanly roll down if the initial pre-flare is good. The hand-press worked too with this method, though it took some manual pre-swaging before it worked well. Finally, today I picked up a larger diameter center punch (1/4") since the previous one could nearly fit inside the eyelet and now swaging just involves one good thwack of the hammer on the center punch to pre-flare then sticking the board in my hand-press and cranking until the roll is finished but the eyelet is not squashed. Nice! Picture below of a recent eyelet using the new process:
   
Perhaps this isn't that interesting for those of you who got the london power circuit card kit with instructions, however I bought the (presumably comparable) components separately off digikey to save on shipping from Canada so I've had to experiment with things myself to figure out the process. I've probably burned through around $3-$6 in eyelets by now, plus much more in time, so perhaps I didn't save money after all.

Anyhow, here are pictures of the test fit for the bridge. I have formed the leads but not trimmed them yet, so the diodes won't actually stick up that far. In hindsight I should have left more space to work in, once I add the wires it will be a tight squeeze working with the soldering iron 'n stuff.
   
   
Reply
#30
Hi Guys

The 1715 staking tool is the correct one for $44 Keystone eyelets. There are three ways to use it. In all cases, the eyelet is inserted into a 1/8" (3mm) hole with the large roll on the top side of the card. The card is placed upside down on a hard metal surface - not wood or plastic.

1 - Hammer: Holding the tool perfectly vertical over the eyelet will almost always result in a split eyelet. You have to gently swivel and roll the end of the tool over the rim of the eyelet while GENTLY tapping it with the hammer. The eyelet is small and the hammer should be relatively light itself.

2 - Drill press: Secure the tool in the chuck of the drill press. Use the chuck key in more than one of the tightening holes to assure that the tool is locked in as well as possible. Shift the drill press table so the eyelet to be pressed is not over the center hole; rather, that it backs onto the solid part of the table. use the drill press lever to lower the tool onto the eyelet and give a surge or two to get more force until the roll is complete. personally, I never found this method to work too well.

3 - Arbor press: This is the best method if you can find a 1-ron press for a reasonable price - and of course, it is difficult to justify this purchase unless you are going to install A LOT of eyelets and/or can use the press for other things. One other use is to swage in captive nuts for chassis covers. The key to using the arbor press properly is to have a secure mounting of the tool. Depending on how large the hole is in the press this will be an issue or not, but it may take some time to get this lined up correctly. I was very lucky to have a customer who is a machinist, who copied the end of the 1715 onto a piece of round-bar that fit snugly into the arbor press. Using the proper size hole for the eyelets, I could load a section of eyelets into the card, flip it just go press press press ... and every roll was perfect and quick.

Regarding tight layouts: While still doing eyelet card construction, my layouts were getting a bit crazy and I had to resort to a smaller eyelet for some sections. Of course, that meant setting up a second arbor press. Inserting components and adding wires had to be planned out so I could work from one end of the card to the other, say, so as not to burn any components.

Tight eyelet layouts are a great reason to use Teflon insulated wire. You do not have to worry about melting the insulation with the soldering iron AND the overall size of the wire is smaller than with PVC insulation.

Fuse holders on eyelet cards: I have a drawer full of the open-style fuse clips as shown in the pics taken by Physics. Mine were for full-size fuses and needed four eyelets. because of the clips themselves, the spacing had to be a bit generous. The pins that go into the eyelets need to be bent prior to soldering to provide mechanical resistance to fuse extraction forces.

Fuse holders designed for PCBs tend to be more compact in some regards, and are mostly fully insulated all around. This allows placing them up against each other with no worries of shorting to each other. The pins only need a single eyelet, which should be placed so the pin can catch the side of the eyelet leaving most of the opening available outside of the fuse holder footprint.

Don't forget to add a bleeder resistor across each of the caps in the voltage doubler. This helps for turn-off safety and also assures voltage sharing.

Have fun
Reply
#31
Thanks for all the info!

I did try using a light hammer and swiveling, and was unable to get a good roll. Sounds like it was user error. Perhaps I was tapping too hard. I was also using a flat patch of cement at that point in time instead of the sheet of 16 gauge steel I picked up a few days ago, so maybe that didn't help either. Either way, I think the hammer method will likely be too time consuming for the amount of projects I have scheduled, so I'll shelve that for latter.

Regarding the arbor press method, I think what I'm currently using is an arbor press. Probably a less than ideal one? It's labeled as a PanaPress, and from my research is likely a 1/4 ton press. There is play in the ram, I can wiggle it enough for the end to move a few millimeters when it's fully extended down to the circuit card. Is that normal, or is that effectively making my tool insecurely mounted? The tool itself is in a 1/4" hole in the end of the ram, though it doesn't sit normal to the ram face last I checked. The hole isn't drilled/reamed deep enough for the tool to sit in very far and it gets narrower at around 1/2" into the ram so the tool can't go all the way in, which I figure is part of why the tool isn't normal to the face. It also means the tool gets wedged snugly inside the hole which makes for a fun time getting it out... I'm the guy who tried drilling the hole out larger so don't blame PanaPress for the poor drill job. I stopped after my drill bits stopped having any visible effect. Perhaps dulled? I kept them flushed with oil, but maybe that wasn't enough. Also, the press doesn't have those fine adjust screws for the ram that I see some other presses have, so I can't remove the play that way.

Using said press with a steel block under the eyelet, the eyelets still came out poorly. Mostly folding over or collapsing. I'm not certain which part of the setup is the issue. Based on what you said I'm thinking either the play in the ram (necessitating a new press? shims?) or the not-quite-normal mounting of the tool (possibly fixable?).

Anyhow, there are some respectably priced used presses near me, including a used Dake, which I hear is a good brand, so I'll I'm going to look into those, see if I can try them out before buying to see if a different press works better. There are some cheap harbor freight presses too, but trusting those seems like a bad idea.

Thanks for mentioning the PCB fuse holders. Initially I didn't see any rated for above 250VAC so I avoided that style, but some more finangling with digikey revealed some rated for 600V(DC?) so that should help with spacing. No multi-circuit high voltage fuse blocks so far though, I'll have to keep looking if I want those instead of individual fuse blocks. I did try to space the eyelets to just catch the edge of the pin this time, but I didn't use calipers to mark things out so they came out a bit too close together as I think the picture shows, with the fuse holder overhanging the eyelet a little. Next I'll use calipers.

Totally forgot about the bleeder resistors when I was assembling this, despite planning for them earlier. Two more things to figure out how to cram into that space, or perhaps install two more eyelets and some jumper wire. Teflon may be a nice idea in the future, I'll have to look into costs.

Thanks for all the info! I'm gonna go try the arbor press only (no hammer & center punch) with the eyelets once more just in case I missed something.
Reply
#32
Update: I just tried some more eyelets while paying close attention, and it's the press' wiggle room causing issues. Compounded by how long the tool is, when I bring down the press the ram shifts all over the place so that it folds over or collapses the eyelet. If I bring the ram all the way forward in it's range of play and then set it atop the eyelet, then it doesn't move front to back, but still does left to right. Maybe if I move it all the way to front left or front right then it won't move at all? Anyhow, if I don't want to deal with that it looks like a new press, or some sort of shim setup, may be in order.
Reply


Forum Jump:

[-]
Come in where it's warm!
A warm welcome to tube amp modding fans and those interested in hi-fi audio! Readers of Kevin O'Connor's The Ultimate Tone (TUT) book series form a part of our population. Kevin O'Connor is the creator of the popular Power Scaling methodology for amplifiers.
Please remember these three principles: respect, sharing, community.
Not familiar with The Ultimate Tone book series? See discussion topics, or click here to visit London Power/Power Press Publishing.

[-]
Tube Amp Forum Hosted by London Power
London Power logo