Custom Chassis?

All good stuff ... this keeps up, I'll have the knowledge to put a meter on most anything, including my toaster.  

Meter circuit v3.2 (beta)

Didn't like the mV meters ... they're made to mount behind the panel and aren't going to work unless I add an extra bezel and space that out another half inch up from the main panel. Yuk. So ... back to one of the original ideas of using the original mA meters, but converting them to mV ...

Snip!



Easy so far ... the clear covers pop right off and two screws to remove the scale. Just bent the shunt resistor lead out of the way and that's the only internal mod. I went with a VR on the plus terminal instead to adjust the meter to match the circuit. They're log scale, and I'll have them turned all the way down prior to powering up the first time.



Had some nice 50k laying about - those might be a bit twitchy to adjust, but no worries about not having enough range to work with. Mounting them was simple enough - soldered them in, then hot glued to the back of the meter is all.

NOTE >> There IS another resistor already in place on the hot lead ... 0.75 ohm ... I left that in place for now. The external VR should still be able to match the meter, unless I end up with a negative value. Those HopSun meters actually have quite a bit of room to work in, so be easy enough to remove if it's an issue.

That's as far as I got last night - it was getting late, and I want to fire it up without the meter wiring in place just to make sure things are good as is. All that's left then is to wire them into the amp. The amp's been rewired to the original schematic otherwise. The meter feeds will just be piggybacked onto the same test points I use for the VOM readings and be selectable with the SPDT mini switches which now have a center off position.

Short version ... I'm pretty much back where I was, only with the modded meters w/VR and the center off switches ...

Anyway ... that will leave me with this as the final(?) circuit ...

Huh,

Shows how much I pay attention to the pictures.   Anyhoo... glad to be of service.

Progress report ... got the wiring done and the distortion is gone.

However ... the meters still aren't right.

I can match the VOM test points and the meter using the 50K VRs on the back of the meters. It's very twitchy as I expected ... have to head over to ratshack for a couple 10k linear pots maybe. Problem though ... the reading at the test points drops by 4mV when I switch the meter out of the circuit. Drat!

Thing is, with the meter switch off, the meter itself is completely out of the circuit. Would the wire lengths be the culprit? Those are twisted 22 gauge, approx a foot long. Also wondering now about that internal 0.75 ohm resistor on the plus lead ...

On the bright side, the amp's working fine now, so I disconnected the meter circuit from the test points once again to eliminate any error potential when setting the bias, but I'd really like to eventually track this down.

dOH! One thing I forgot to check is whether the 4mV drop is at all the test points whenever a meter is switched on.

I think that 4mv difference is the difference across your 0.75 ohm meter resistor

Seems reasonable, As they say, I had a concern that would be the case.

It's been pointed out that what I should be doing is adjusting the VOM test points for the proper bias with the meters ON, and just leaving them on. That also eliminates having to pop the internal resistor in the meter that's in series with the plus post. That looks tricky, and my hands aren't as steady as they once were. The center off thing was to eliminate the meters pegging the stops under load, and that's no longer an issue. Now, using the original piggyback circuit, with the converted meters and VRs to adjust the range, I'm only getting a point or two variation when the music's playing, so that will work.

The current 50k VRs I'm using to adjust the meters are ... tweaky. VERY narrow range from stop to stop. Made a trip to the local ham/cb radio repair, said howdy and 73's, and scored a couple terminal strips and four 22k VRs. I plan to use one VR per tube now so I can match some variance due to the wire lengths to each test point (keeping in mind mine are a LOT longer than with the standard dynaco chassis). If the 22k pots are still tweaky, I got some 20k resistors to bridge them and take them down to around 10k range. I should be able to get an exact match to each test point this way with a lot more turn to set the range.

I suppose I should also replace the center off toggles with the original On/On SPDTs I was using to eliminate the lag when selecting tubes. That will keep the circuit resistance constant, with a slight peak when switching tubes in a bank. Not a biggie, as we're only talking 4mV when the switch breaks ...

Getting close ...

Right now, I got it back in the stereo room and am just groovin' on da muzak ... this thing is ... what's the technical term ... Freakin' Amazin' !!

Well ... finally ... got the beast on the bench and finished up on the meter circuit. As expected, the only issue was the rating on the pots used to adjust the meters. Rather than fiddle fart around substituting values, I got one of these off da bay ...



Fun little toy that should be a handy addition to any hobby bench! Dialed it right in at around 100 ohm (which explains why 50k was a bit ... twitchy) and went with 200 ohm pots, one per tube. I hot glued a couple scraps of perf board to the back of the meters to hold the pots and give secure mounting for the wires. Wires that came up a bit short with all the variations of the circuit I tried - I'm tempted to rip the whole harness off and redo it so I can bundle the wires nicely, but ... must ... resist ...



Next came dialing them in ...

- Adjust the bias to 55mA at the test points with the VOM, and with the meters off
- Adjust each trim pot on the meters to 55mA

So far they're reading consistently. I do have to switch the meters off when I'm actually setting the bias at the test points as those read high due to the additional resistance with the meters in line, but that's where the center off SPDT switches are nice. I like being able to just give the meters a glance with a couple quick flips to make sure the tubes are all working correctly and that they're warmed up. I'll play around with tweaking the bias just using the mark one eyeball and verify with the VOM to see how close I can get.

~ ~ ~

Speaking of tubes working correctly ... It's the little things ...

When I was done playing about with the meter circuit, somewhere between the basement shop and the audio rack something went horribly wrong. V6 started red plating on power up. Huh??

Figured I fat fingered something when I had it on the bench, so ...

- shuffled tubes - the problem stayed at V6
- checked voltages (all good)
- checked continuity to eliminate the chance of a stretched wire. I've seen stranger things happen.
- checked bias resistor (tested good, but I replaced it anyway)
- double checked all solder connections at V6 and the VTA board, and retouched a few here and there.
- gave it a good check for whiskers and wire cuttings that could be shorting stuff out.

Finally ... noticed the bias pot for V6 had a sticky spot ... kinda raspy and tested open. Replaced that, and there was once again joy and jubilation here in sKiZoLand ... as for the rest of the testing, hey ... I needed the practice anyway.

Moral of THAT story ... Last time I buy cheap Chinese junk pots of da bay. I plan to replace all four with some decent ones. Anybody know a good source? I found these roller pots over at All Parts for $6 a pop that look like they'll do nicely.

http://www.allparts.com/EP-4992-000-50K-Linear-Roller-Pot_p_1486.html

UPDATE >

Should have gone with my first instincts. I went with some OEM Fender 50k pots that were a couple bucks less each, and come to find out, they're NOT a linear taper. Halfway up and they're only at around 10mA and only around 20mA at 2 o'clock. From there, it get's tight and another step or two puts me at 55mA, but it's a bit twitchy because the range tightens up so much. MUCH better pot though. Ah well ... I may order up the roller pots from All Parts anyway, and hold them in reserve for the next time I go under the hood.

Here it is with the Phillips 5R4GYS and TungSol KT120's installed.



Plenty of room for them, and the board biased up nice with plenty of extra range to play with. Didna have to mess with the "mystery resistor" at R39. The amp doesn't seem to be working any harder than it did with the standard tube set. I could still hold my finger to the power transformer after a good hour of hard running. I plan to keep it at 55mA until the tubes break in, then eventually run them up to 60mA and leave them there.  

This is also with a set of 12BH7's as channel driver ... not sure if that adds any significant load, but that extra overhead is coming in right handy here! Thanx, Bob!


Oh. PS ... the meters work!!  

We all can tell that MANY hours went into the planning and labor involved in building the amp.

All true, but can't stress enough the importance of pure dumb luck and a lot of help from you and the board ... also worth mentioning ... many hours of FUN went into this thing, and many more to come!

One last thing I'm gonna mull over is how to close up the bottom. I was just gonna go with a perf grill or expanded metal of some kind, but tripped over a laptop cooler the other day in the basement.



Heavy metal and low profile, with an 8" fan on it and is dead quiet. Low rpm's too, so it's more of a circulation kinda thang. Size is about right too, with a couple filler plates on either end. It would tuck in nice and be invisible once mounted, and plenty of vents up top around the tubes. Wondering if I could tap 5v off the amp, but I'd probably be better off just plugging it into the wall with a USB wall wart ... only real concern would be pulling dust into the chassis. Even if I decide against the fan, the metal piece will make a nice bottom panel.

And ya ... I know what you mean about tube rolling ... I can see where that could get crazy AND expensive over time. Wanna have some fun, count up how many tubes you got and plug the numbers in here ...

http://www.mathsisfun.com/combinatorics/combinations-permutations-calculator.html

I'm already up to 455 possible combinations here.  ... I better get to rolling!

sKiZo wrote:

Are you a car guy, too? Is the badge a tribute to a car you owned long ago, a mid-1970s Mercury Monarch? It looks
like the dashboard badge from one of those models.

Give the man a ceegar! Still got a few other parts laying about, but figured maybe the grill would be a bit much on this build.

PS ... the Limited sticker comes from a golf cart ... cost $4 on da bay ... couldn't resist ...

http://www.ebay.com/itm/LIMITED-EDITION-CUSTOM-DASH-BODY-EMBLEM-CARS-GOLF-LE1-/161105628483?hash=item2582a4cd43&vxp=mtr

You just HAVE to figure out a build where you can conceal the user controls behind a Mercury Monarch gas cap door.  Only problem was, they tended to break.  Most owners who had one of those cars for 100,000 miles went through two or three of the darn things!

Also you'd have to add lettering to the left and right of the gas door.  Otherwise it'd say only
"NAR."  No fair substituting one from a Ford, which has the entire script, "Granada" on it!

I grabbed the photo off the web, but out here in rust-free California you can still find mid-70s
cars on the road, including Ford Granadas and Mercury Monarchs. In fact, this one is from
GILROY, CA.

OK then ... rainy day, so I went and figured out the total damages ...

VTA ST-120 tube amplifier kit                      $1,015.00 (includes two spare bottle rectifiers)
TungSol KT120 (matched set of four)              $201.75
Phillips 5R4GYS rectifier                                $48.00
RCA 12AU7 clear tops                                   $30.00
WS1 copper top rectifier                                $17.00
sylvania 12BH7a  (pair)                                    $7.91
triad C17x chokes (pair)                                 $38.71
ST120 chassis top plate                               $120.88
ST120 chassis back plate                              $39.49
Makore for amp case                                     $18.00
trim pots/resistors                                         $50.00
Velleman meters                                           $27.90
power switch                                                  $8.38
limited edition nameplate                                 $4.49
Paint & Varnish                                             $30.00
Misc parts & shop supplies                             $80.00
                                                          $1,737.51

The fun and experience ... priceless!

I've got a couple hundred more into spare tubes, but all in all, I figure I could have got a lot less for a lot more ...

The amp turned out very good! Nice find on FPE, im gonna see about having them build my panels.

Question, is it best to get ma or mv meters? I will have to go back and find the post but Bob mentioned reading mv and them equating to ma or something to that effect....?

Thanks....again, nice job....has a nice retro look to it....kinda reminds me of the heathkit Monoblocks, cant remember the model offhand.

You can go either way on the meters, mV or mA, but you would need to convert an mA meter to read mV using the same circuit I used. The conversion consists of opening up the meter body and clipping the shunt resistor, and adding an external trimmer to match the meter to the bias test point reading. That's all discussed earlier, so I won't bore you with that.

Main reason I went with the mA meters is that decent mV meters in this range are almost impossible to find ... at least ones that I liked. Maybe you'll have better luck and save a step or two in your build. I suppose you could even use VU meters, but I didn't get into that so you're on your own there. I really like the Velleman's as they have a large easy to read scale. As is, I don't get too fancy on calibration - the meters have a nice mark just below 60mA that I've designated as 55mA.



The trim pots for each power tube allow me to put that pretty much anywhere I want - I used that mark just because it's handy and easy to eyeball for quick checks. The trim pot adjustment is a one time deal to match the meters to the circuit. You'll still want to have test points where you can get an accurate reading with a VOM every month or so.

And ya ... FPE is killer for getting a nice clean CNC and finish. I saved a couple bucks and allowed for a fudge factor by drilling the small stuff like socket and board mounting screws myself. Best to do all that before actually building the thing.

I popped my FPE design up to the web if you want to take a peek and get your hands dirty. One thing real handy is the octal socket with vents - you can copy and paste that to whatever design you go with. I should load that up to FPE as a "user template" and maybe save some other folk a bit of work. Anyway, feel free to co-opt any part you want for your design - you'll need to download the free FPE software to open the file.

http://www.brainfartz.com/images/Stereo%20Stuff/VTA%20ST-120/st120-build/st120_chassis_skizo.fpd

Also, if I were to do it again, I might increase the holes for the driver tubes to 28mm vice 25mm that I used. The tubes are a bit close to the plate and the extra room would allow for better ventilation and less heat transfer. Not a biggie as the plate gets warm but not hot.

PS ... if you DO try to do any bends, I'd suggest you either use thinner plate stock or have FPE trim the thickness of whatever you use right in the bend area. They added that feature after I built mine. Basically just cutting the thickness on the back side of the panel. My plate was a real bear to bend, and I wasn't going all that far. FPE doesn't do bends, so you'd also need to build your own jig to make that happen, or have a local metal job do it in their bending brake. Also, if you do go with thinner stock, I'd be concerned about the weight of the transformers - you'd probably want to add some bracing in that area.

Good stuff Skizo, thanks!

Btw, I have never seen anyone post this but a tip for bending aluminum is before you bend get a torch and adjust acy so the soot is coming off the flame then run that soot along your brake (bend) line. Next adjustyour torch until normal flames and run your torch along the brake line until soot disapears...the aluminum is now ready to bend and you will not get any cracks. I see no reason a propane torch will not work but this is what we did whenever we had to bend aluminum especially thick diamond plate as standard practice. A lot of it depends on the bend radius and alloy but if you preheat you never get cracks and its easier to bend. I was a metal fabricator for 18 years before becoming maintenace electrician.

Also might add...you dont want to get the torch tip too close to the material and you want fast strokes along the brake line.....dont stay still or in one spot or you will ruin your metal. The key is once the soot is gone the temp is correct. It does not take much.

I will be going with 16ga but will have my transformer covers double as a stifner and cover. Will more than likley have front panels built by fpe although I could and may do it here at work. Its all dependent on time as I also have remodeling going on at home as well as planning our home theater. Gotta save where I can with both time and $$ but in some instances forced to shell out some green..... ugh.

Hey! You need to readjust your priorities!! Audio FIRST  

Heat can be your friend, but I'd check with FPE first about it's effect on the coatings they use on some of their plate stock. I got the anodized chromated stock (which is more of a matte finish) and I suppose a lot would have to do with how they apply it. The do have raw stock, and powder coated, and Perspex - whatever that is. Seems to me that would be a high gloss plasticoat or some such that wouldn't take to kindly to bending.

I suppose if worse came to worse, you could always strip the coating and just polish it up ...

Stock I used was twice the thickness (250mm) of what you're talking though, so I don't expect it'll be an issue - 16 gauge should bend fairly easily. If you do go with the A/C finish, you'll want to keep a fairly large radius ... I expect it wouldn't take too kindly to a hard 90 ... Original plan was to do that for a back plate, but I decided against that and went with two pieces.

PS ... no sense rushing ... I think it was a good three months after I bought the kit before I actually got going on it. Good thing too, as it didna turn out anything like the original concept.

Yes, stock raw aluminum from the rack (that's how we get it, its painted, powdercoated or anodized after fabrication) ... of course if it is anodized all that heat goes out the window. I'm speaking of raw sheets and thicker material but it also helps lighter material if you do not have the proper tonnage (home made breaks) and a raw finish.... I little bit of heat can help, a little bit, not a lot.

The aluminum sheet metal typically does not need to be preheated unless the bend radius is too tight for the thickness but even that can be dealt with by bending perpendicular to the material grain. However, some of the higher alloy sheet metal will crack and need special attention when fabricating them, even at 16guage is needed. The softer alloys are a piece of cake to bend but harder to weld.

If you buy anodized sheets or if the sheets have the coating to preserve the material from oxidation it changes things, but If by chance you do want to bend a piece that has been anodized it can still be done. Put masking tape on the dies and the material, and when you bend it there will be a separation at the break line due to the material stretching as it is bent but the tape will protect the finish as it slides so you do not scratch the anodized finish...then you run a sharpie or felt pen along the edge of the break and wholla. No scratches and nobody ever knew. Of course the sharpie has to match the color of part that has been anodized. If its been clear coated just mask the edges and shoot a coat of clear at the break line or clear coat the whole chassis. fine sand along the break to blend the area before painting clear coat.

Another thing to ask for if you plan on doing a bend at home and the material is a little thick, request that the grain runs along the longest break line. Its easier to bend when the die and the grain of the material are parallel but...its also easier to crack depending on the alloy. Million tricks of the trade but I'm just throwing a few out there...
...for anyone who may need the info some day...


FTR, I will be buying my chassis already built :Dbut I will do a few extra mods to them. There is an outside chance I may build them at work because we have all the tools in our metal shop but when the horn sounds and its time to go home I would rather be home instead of stuck at work voluntarily, lol.... no place like home!! :DSo there is my dilemma.

Honestly you really did a nice job on your chassis. I'm just so darn used to having everything all metal its just the way I think. I gotta think outside the box! I can do some woodwork at home.....

ah, yes priorities..... let the music play!

I suspect I will be going through a few design changes too before the end result ...heck I already have