|
|
Post by svart on May 20, 2014 13:17:10 GMT -6
So I've started down this path a bit. I already have a set of I/O transformers and most of the parts needed to build the unit leftover from my 1272 builds.
There are a few 2254/33609 hybrids floating around, mainly from the days of Igor, but I didn't care for the changes he made. I decided that the best course of action would be to roll my own boards with my own changes and then I could play around with it.
So over the course of a number of weeks I captured the schematic of the 2254 signal path with a 33609 sidechain and studied how things work. Honestly it's pretty simple overall but complex in the implementation. I expect there to be many problems along the way getting things working together.
I've already gathered the 1x24, 2x6 and 1x12 rotary switches from various places. The 1x24 are Russian and don't have threads for mounting. I'll be threading the body around the shaft and running it right into the front panel. They are pretty big too, so there will be fitment issues I'm sure. I'm certainly not paying 50$ each for elma switches.
I've already submitted the PCB design to the PCB maker and I'm now waiting for the boards to show up. I decided I would have them professionally made with plated through-holes due to the possibility of a lot of soldering. I also wanted to try the "budget" PCB service and see what kind of quality you get for the prices. I've also noticed that a number of photoetch PCBs have had issues with exposure. I'll have one board expose just fine, and the next one won't, even with the same lamp and same timing. Not sure what is going on but if I'm going to spend a bunch of junk etches, then I might as well get the real thing done.
I'm attempting to see what I can get away with budget wise. My expectation is that I can get this done for 700$ or less for a stereo unit with the transformers and chassis being the most costly items. Therefor I'll be using more budget minded parts, like transformers and meters from Hairball, a DIY chassis and probably FPD front panel where I drill and cut my own holes.
Anyway, I'll keep this updated with what happens.
|
|
|
|
Post by svart on May 22, 2014 8:03:04 GMT -6
Supposedly the boards will be delivered today. I guess that means I might start doing some stuffing tonight if all else goes well.
|
|
|
|
Post by svart on May 22, 2014 21:25:12 GMT -6
So here's my board. www.facebook.com/photo.php?fbid=10152130205660679&set=a.10150936019350679.419561.699740678&type=1&theater Looks pretty good except I forgot to resize the holes for some of the larger leaded parts. I'll have to drill those holes and just rely on the pads on those parts. I also bought some trimmer pots on closeout over at Newark/Farnell but I didn't realize the spacing was different, so even though I have dual footprints on the PCB for the trimmers, these fit neither of them. I'm using BC560C in place of all the BC214C and I think I might use MPSA18's in place of the BC184. I do have some BC184 I could use but I don't know if I have enough to populate both boards. I've also read that replacing the BJTs doesn't make much difference in the sound and that a lot of repair shops are using whatever they have on hand now that BC184 is gone. |
|
|
|
Post by svart on May 23, 2014 7:24:51 GMT -6
Looks like MPSA18 for the BC184, 2N3053 for the BC441 and BD438 or MJE371 for the JE371.
|
|
|
|
Post by svart on May 23, 2014 8:52:34 GMT -6
Trying to figure out if I can use the 10468 as the input transformer as well, just reversed.
Also trying to figure out if the polar caps in the sidechain are electrolytic or tantalum. Pics are hard to see, but look like 'lytics in the sidechain.
And shoot, it looks like I have to mount the MPSA18 and BC560 reversed compared to the BC184 and BC214. Oh well, at least it's a simple change.
BC547 looks OK but would need to be sorted for HFE.
|
|
|
|
Post by svart on May 23, 2014 21:42:55 GMT -6
Tonight's work.  Turns out the wrong trimmers really screwed me. I can't even bend their leads enough to mount them. Guess I have to order a handful more. Each board is a single channel and power supply. I'm only going to stuff one power supply and then just jumper over the output from the other one. The company I used was 4PCB. They have a 33$ per board special for 2 layer boards under 64sq inches and under a certain amount of holes and some other restrictions. Minimum 4pcs though. I started a front panel. I know there are some available out there in the world but I plan on using certain meters and having certain options for the ranges and things that aren't on the kits that those front panels are made for, so I have to make my own. Unfortunately the costs have gone way up for front panels it seems. Almost 100$ for what I want. Sheesh. I remember getting them for less than 50$ maybe 6 years ago. |
|
|
|
Post by Calvin on May 23, 2014 23:06:35 GMT -6
This is a great project. I'm looking forward to seeing your progress.
|
|
|
|
Post by svart on May 24, 2014 23:49:40 GMT -6
So i got a lot done tonight. I got audio through the ba283 portion and it sounds good. I also got audio through the front end after messing with the biasing on the front end fets. Not sure why i had to do that. It may be because I'm using different pnp transistors and the biasing needs tweaking.
I also found that while the mpsa18 and bc560 pinouts are backwards, so are my symbols, so they actually go in correctly! The bd438 pinout was different on this part than what was in my schematic program, so i had to swap pins on it.
Overall, i can make the audio drop maybe 10db by swinging around the CV bias trimmer, which also means that the diode bridge works close to how it should.
It was a good day, i got further that expected and with fewer clusterfucks than i expected too. I have to wire up all the side chain controls, which i didn't expect to be doing so soon,before i can go further. I suspect the cv signal also biases the fet gates some, and that's why i had to goose the front end to work.
|
|
|
|
Post by svart on May 25, 2014 0:03:22 GMT -6
The hairball transformers work fine too. The front end definitely likes the 10468 turned around backwards. Standard installation gives higher levels but sounds strange. No need to figure out why. The ba283 section has a few trimmers for gain of the two sections. I might end up taking them out and wiring it just like a 1272. I also have the volume and trim pots just like a hotrod 1272. The trimmers go from like normal to distorted loud in a turn or so, and these are 28 turn trimmers. Go figure.
Now i need to figure out what to do with the other two boards.
|
|
|
|
Post by jcoutu1 on May 25, 2014 6:30:24 GMT -6
The hairball transformers work fine too. The front end definitely likes the 10468 turned around backwards. Standard installation gives higher levels but sounds strange. No need to figure out why. The ba283 section has a few trimmers for gain of the two sections. I might end up taking them out and wiring it just like a 1272. I also have the volume and trim pots just like a hotrod 1272. The trimmers go from like normal to distorted loud in a turn or so, and these are 28 turn trimmers. Go figure. Now i need to figure out what to do with the other two boards. Build me one?  Pm me a price if you're interested. |
|
|
|
Post by svart on May 25, 2014 8:02:43 GMT -6
Well I can tell you it won't be cheap. I might release the gerber files for the boards and/or files so people could etch their own, but I don't think I'll be making these for kit builders or making units to sell. The way I did the layout is not conducive for good production, it was for ease of prototyping. I still might make a board to span the backs of the front panel switches, to avoid many hours and many wires.
|
|
|
|
Post by svart on May 25, 2014 8:43:20 GMT -6
decided that I'll hook the amp sections up more like a hotrod 1272. I don't need all the trimmer adjustments.
I may also just hardwire the "trim" between the ba283 sections. That way there is only a volume adjustment coming out of the compression section into the makeup gain section. Not ideal for noise, but ideal for fewer unnecessary controls.
|
|
|
|
Post by svart on May 25, 2014 13:58:51 GMT -6
 So for today, I hooked up some pots that approximated the values of the rotary switches and I now have compression. It turns out my two front BC560 were backwards on the schematic.. My bad. The rest are OK. I thought I had to swap pins on the BD438 as well but I think that was the sleepiness talking last night and it was OK as is. So for today I got audio through it completely and got some degree of compression. I still haven't hooked up the limiting yet, but I think it's going to be a matter of needing to put this and the other channel into a chassis to deal with the wiring from this point on. It's getting a little unwieldy with all these things hanging off the board. I'm also thinking of using dual concentric pots for things like the volume and gain trim to save some space. Unsure yet though as finding concentric knobs is hard to do. If I can't get the 24 pole switches to fit, I now know that pots configured as rheostats will work OK. It's just getting the values right. I know these things aren't really meant to compress like crazy, as they only offer 6:1 at most, but I noticed it get's really grainy when smashing the signal. It's unknown yet if that is a calibration issue that will go away after calibration, or if it's a characteristic of diode compression. On second thought, if I had to do this over, I'd probably just make a stereo unit on one board. I could have fit both L and R amps and a single sidechain and just summed the feeds like what was done on the old GSSL. This would have saved a lot of time and space, and probably could have made it a 1U device. www.musictoolz.pl/pliki/33609j-pdf-77.pdf |
|
|
|
Post by svart on May 25, 2014 14:30:16 GMT -6
I also noticed that my variable attack mod does indeed work. It's a play on the fast/slow attack switch of the limiter section, which varied the amount of current to an RC timing section. It might take some playing around to get the timing constant figured out and adjusted.
|
|
|
|
Post by svart on May 25, 2014 23:14:09 GMT -6
 I built the ratio and compression threshold switches. Both work as intended. Still worried these russian switches are too big. I might have to stagger them on the front panel. This is going way faster than expected. I suppose I need to order the other transformers and the chassis/face. |
|
|
|
Post by jimwilliams on May 27, 2014 9:46:19 GMT -6
Try Japanese bipolar transistors, they are less noise. 2SA1084 Renesas PNP's are .6 nv. Matching 2SC2546 NPN's or the Toshiba version, 2SC3329's are excellent.
I used those in my GAP 73 and it's very quiet now.
|
|
|
|
Post by svart on May 27, 2014 10:12:08 GMT -6
Thanks Jim. I'll try those in subsequent builds. These seem to work fine so far though. This was nothing short of an experiment to start. Maybe more tweaking will come of it later. The MPSA18 are pretty high beta/HFE, all of the ones I measured were around 950-980.. So they start off much better than the BC184 that needed to be selected for >600.
|
|
|
|
Post by rocinante on May 28, 2014 1:20:53 GMT -6
Wow. I can't believe how fast you are moving along with this. The 'great gods of outboard hardware' will be bowing to you in no time. Very impressive.
|
|
|
|
Post by svart on May 28, 2014 6:49:21 GMT -6
Thanks. It's been going faster than expected. I took a long time getting things straight during the layout process and did it in a way that makes it easy to work on, like putting all the recovery circuit on the board and just wiring the switches over to it. Having to wire all that onto the switch would be a nightmare I think. It makes more wires, but those are easy!
Anyway, last night I wired up the recovery switches for the compressor and limiter. I didn't test them out yet though. I still don't have the limiter in-circuit yet. I still need to do the level switch for it. The next couple days I should have that done I think.
I also started on building the switches for the second channel, although I haven't even powered that one on yet! I think I'll buy a rack chassis today so I can put the boards in it while waiting for the front panel. I've done about half of the front panel and have yet to decide if I want it more pretty or more functional. The cost can be considerable it seems.
|
|
|
|
Post by svart on May 28, 2014 13:59:23 GMT -6
Mostly done with the front panel. I haven't done anything spectacular as I'm trying to keep costs down. I refigured the scales to fit the degrees of the switch positions. I think they are right but I'll do a print out and match it up for real before ordering the front panel. Right now it sits at around 80$, which is more than I expected it to be since I have more ornate front panels from years ago that were considerably cheaper. I still need to decide how to mount the russian rotary switches. The will not sit over/under in a 2U box without scrubbing the top and bottom lids pretty hard. I'm thinking about offsetting them but when I do this, the front panel looks strange.
|
|
|
|
Post by tonycamphd on May 28, 2014 20:20:46 GMT -6
Mostly done with the front panel. I haven't done anything spectacular as I'm trying to keep costs down. I refigured the scales to fit the degrees of the switch positions. I think they are right but I'll do a print out and match it up for real before ordering the front panel. Right now it sits at around 80$, which is more than I expected it to be since I have more ornate front panels from years ago that were considerably cheaper. I still need to decide how to mount the russian rotary switches. The will not sit over/under in a 2U box without scrubbing the top and bottom lids pretty hard. I'm thinking about offsetting them but when I do this, the front panel looks strange. Awesome Svart! i say offset them, i like strange, and it puts you in a league by yourself, which you pretty much are anyway, great job on this man 8) |
|
|
|
Post by svart on May 28, 2014 20:25:27 GMT -6
I'll have a few more pics later, maybe tomorrow.
Ordered the chassis a few minutes ago. Now I just need to get this faceplate finalized and sent off.
|
|
|
|
Post by rocinante on May 29, 2014 1:11:38 GMT -6
I really wish there was a more affordable option other than fpe. I have three builds right now that I will eventually send off; like the design is done, matched, finalized and I even made a temporary frontpanel using transfers with until then. Needless to say I don't love the front panels and think they deserve a much better (more attractive) face. I also have three more to do (Neumann w492, another 1176, and a 1292,) that make me cringe when I think about all the frustrations that come with front panel drilling and layouts. And at anywhere from $70-$110 a pop that's pretty expensive. I've been trying to contact trophy shops recently in hopes of finding a better alternative. If I find one ill definitely spread the word.
Incredibly interesting build diary BTW.
|
|
|
|
Post by svart on May 29, 2014 7:43:13 GMT -6
I don't know if it's still available but I used Chae Ham's front panel service for my Sontec clone. The job itself was cheap and it looked very nice (laser etch on smooth powder coating) but I had to buy and ship him the panel. All said and done it was around the same price with all the shipping included. I do appreciate that FPD offers the software to do the front panels, and I'm willing to pay a little more for the convenience of that. Anyway, here's the pic of the day:  I'm contemplating just going with a Fast/Slow switch for the compressor Attack rather than a pot. Mainly this is because I don't have much more room on the front panel! Honestly if it's like any of my other compressors, I'll either use it as it was designed, or turn the attack to some arbitrary long time. |
|
|
|
Post by svart on May 30, 2014 8:42:04 GMT -6
Didn't really do much on this last night. I wired the recovery switches for the other channel and I got some backordered transistors for it and installed them. Otherwise, I'm waiting for the chassis so I can mount the two boards and also trying to finish up the front panel.
Just ordered the other two 10468 and two meters from hairball.
So far, the running total is:
75$ 2x PCBs (150$ for 4 was minimum order)
75$ chassis
50$ misc parts and switches (most other parts I already had)
175$ for transformers and meters
80$ front panel(so far)
455$ out of pocket.
If I bought all new it'd be more like:
75 PCBs
75 Chassis
80 Front panel
125 parts/switches/pieces
310 transformers (hairball)
30$ power tranny
695$
|
|
