TAC Scorpion Overhaul Odyssey

[Potato Cakes]

I had mentioned before that I was going to redo the gain setting of the master faders to have the top of the throw to be unity gain. If I was using using the ICs as the master bus summing amps, then according to Mr. Williams using 22k1 for the feedback resistors for the second half of U11 and U12 (R56 and 57) then this would give me the desired gain setting for the master faders. I then decided to apply this to all of the faders as this would allow me to better recall mixes. For mixing tracks into stems, I almost never have the faders over unity. When I go back and mix down stems, I have been keeping faders at unity and using level control in the software. I record everything at 24 bit so I don't seem to notice any sonic degradation from bit reduction that may take place. Doing this will mitigate some of the inconvenience that comes with mixing down with analog gear and I know if all the faders are all the way up then they're at unity and I don't have to run tone through all of the channels and verify signal at the output before mixing a project.

So that being said, my initial thought was to use this same value for the bus, aux, and channel faders to achieve the same thing. But looking the schematic closer I see that the original value for the fader makeup op amp feedback resistor on the master section is 68k, the aux is 100k, and the buss and channels are 22k. I could just wire in a trimmer at that spot to find this value, but am wanting to know how to figure this out mathematically and to avoid adding any additional soldering to the board as again these PCBs are not made to withstand a lot of tinkering. Essentially I am wanting to reduce the output signal by 10dB on all of the op amps after the fader. The op amps being used at this spot is LME49720NA and all of the faders are 10k.

Thanks!

Paul

 

[Potato Cakes]

I finished adding the ceramic bypass caps to the channels. I came up with a much cleaner scheme for soldering them in place that minimizes the distance from IC VDC connection to capacitor and negates the need for shrink tubing to isolate the lead from other solder points. A similar approach was used for the master, bus, and aux channels.

Thanks!

Paul

 

[Potato Cakes]

Had a bit of a set back today when I fired up the console as I managed to blow some 10uF caps in the PSU. I replaced to those caps and turned it on without the console plugged in and it was fine. So there was something going on from the power cable to the desk. The only things that were installed were the master channel and the ACA-Bo. I pulled apart the power cable to double check it and I found something that I had never done in the thousands of electrical connectors I have wired... I had managed to swap the ground and +VDC rail. I know when I wired it I had checked the markings on the connector right before I mated it with the wire. I'm still quite a bit mad and completely disbelieving that I could have done something like that. I got that fixed then reconnected it to the console and the power was then fine. None of the caps were blown nor were any of the resistors burned on the ACA or the master channel so I moved forward with testing.

This is the first time this console has been powered up since I had acquired it, so the previous state of it's functionality was unknown. I started with only one input channel installed in the console. With it just powered on, the led segments on the meters for buss 6 and 7 were alternated on/off. Buss 10 metering was maxed out. By directly sending signal to the meters from the exposed back panel I verified that all of the other meters were working. When signal was applied to the buss 7 card the other LEDs lit up but not for buss 6. Possibly I mangled the LED driver with the power issue or it might have already been like that. PFL LEDs on the master channel lit up when PFL was selected on the channel. Mute LED on master channel was functioning.

The first test was signal in the channel and back out of the direct outs. This was working fine and I noticed that the direct out was post fade, so at least I know the channel was passing signal. But when I assigned it to Stereo I could not get signal to the output XLRs. With my unbelievably stupid mistake with the power cable, I thought that I destroyed the IC's in the Master channel, so I replaced all of those but still no output signal. I was getting metering but only on the right side. I know the meters work as I had previously verified them. The ACA and the DOAs all seem fine when directly injecting signal at the board.

I installed a bus and aux send channel strip to see if I can get signal from the input channel that seems to work fine, so I'm back to the master bus.

Skipping over all the other boring troubleshooting stuff, this where I am currently. When I pull the master channel, I get signal from the input channels on buss connector pins 3 and 4 that are L/R per the schematic. When the master channel is connected the signal on those pins is killed. On the master buss schematic there is a slate mixer connection on pins 3 and 4 which has the appropriate 10k resistors on it. The thing that is puzzling me is that I can't find on the schematics that I have where signal on these pins connects to pins 31 and 32 which feed the summing buss op amps. None of the channels strips connect to these pins so I have to find how 3 and 4 get to 31 and 32. I'm going to replace all of the diodes on the Master section as well and if that doesn't do anything helpful then I guess it's just going to be staring at the PCB until the answer presents itself. This whole ordeal has been quite a bummer as I thought I would be finishing up this evening.

Thanks!

Paul

 

[Potato Cakes]

WELL....

Almost as embarrassing as the horrible power cable job is the getting two of the ribbon cables from the back panel to the master section backwards. I had labeled the ribbon cables themselves but when putting them back on the back panel PCBs I had the outside ones reversed. Amazing. I'm not sure what is going on the past two weeks with myself and electronics, but it has seemed that my stupid mistakes have increased ten fold.

Anyway, the last thing I need to sort out getting the L/R meters to work when the stereo bus is selected. I had removed the components from the Mix Buss L/R and I am going from the ACA-Bo directly to the XLRs, so I need a jumper that goes back to the master buss channel for the convenience of using the meter trimmers. Then I can finally start testing and put this guy back together for good.

Thanks!

Paul

 

[Potato Cakes]

Again, what I thought would be super simple as not as such. I realized in the process of hooking up a return for the meters that I wasn't going to be getting the full signal by just using one side of each Booster output connection, so I needed a way to get both + and - to the metering section. If I had sent and unbalanced signal from the booster back the master channel and then impedance balanced the main outputs, then everything would have worked fine. But I had opted to wired the booster outputs directly to the main out connectors, so more things had to happen before the meters would work. To prevent from unbalancing the outputs to get proper signal back the master channel for monitoring and metering, I added a pair of simple differential amp PCBs that I happened to have laying around. I used INA134PA's to perform the unbalancing procedure.

 

[Potato Cakes]

Now we have meters and main outputs working. I'm spend the rest of the evening doing tests and changing the fader gain on all the channels.

Soon I'll have to order cables and connectors and start making DSUB to XLR fan outs.

Thanks!

Paul

 

[Potato Cakes]

All the channels have had their gain restructured and now I am awaiting some caps to come in tomorrow and finish the input channels. I'm short a couple of chips so one more order is needed. Also tomorrow I start running tests for noise and frequency response. Hopefully I will be very pleased and not have wasted a bunch of time and money.

Also of note, the design for the 553  daughter cards is complete and I just need to get some prototypes ordered to verify it works and sounds good.

Thanks!

Paul

 

[Potato Cakes]

After some unexpected distractions and procrastination, I finally finished up cleaning all the channels, replacing the last several broken pot shafts, and got it all back in minus one of the bus channels that I had sent with the guy who designed the replacement bus channel EQ. I had to do a little trouble shooting as I had a missed solder joint, a cold solder joint, a missed broken solder pad, and a couple of missing op amps. The only thing I have left one of the input channels is about -6dB from the rest of them. That normally means there is a bad cable as the input being used is balanced, but when I swapped channels there was no longer a level difference, so I have some more poking around to do. Other than that all of the channels are working and there was no issue with the original ribbon cables. I did some noise tests and with a 1kHz tone at 0dBu (-18dBFS reference) I am getting about 90dB S/N with all 24 channels, 2 aux returns, and 15 bus channels set to unity gain and all assigned to the stereo out, which is pretty fantastic for an old desk with this size frame.

 

[Potato Cakes]

Measurement with no signal

 

[Potato Cakes]

I did notice that with the DOA input cards I'm using reverses the polarity at the output of the DOA. I thought it was something that was going on with the ACA-Bo I installed, but the signal going in and out of the ACA part is unbalanced as is the input to the booster section. I'm then going straight to the output XLRs from the transformer. I checked the aux inputs and the tape inputs and those are all are 0 degrees phase with the reference signal. The direct outs from the channels themselves are also 180 out of phase like the main L/R output. I did check all of my cables before testing just to be sure. I just keep the polarity reverse button pressed in all the channels that will solve that.

I discovered that the pan law in the this desk seems to be about 4.5dB, with it increasing in level when panned to the sides. I seems a little high, but I'm not familiar with console designs during this era.

I also didn't seem to notice a difference in level, frequency response, or noise floor difference between the channel I had added the electrolytic IC bypass caps and one that just had the ceramics. Again, all the channels are changed to unity gain only with the DOA at the input, so maybe for my application it wouldn't matter much. If I had kept the stock mic preamp and made tweaks to that then I could see there being a possible benefit with the SSM2015 chip. I may still do some more listening tests with the EQ but I think for my purposes just using ceramic bypass caps will be great.

Now I'm waiting on getting my prototype boards to do some testing with the buss EQ section. But in the meantime I'm super thrilled its all up and running finally. Tomorrow I'm going to finish fastening the channels in place and have another look at the one input module that has -6dB output. I got some sexy screws to replace the rusted M3 ones to help give it some aesthetic appeal. Then it's time to make DSUB fanouts for the patch bays.

Thanks!

Paul

 

[Potato Cakes]

Found the -6dB issue. Apparently I was a little too fast removing solder durning the recap process that I unsoldered the jumper with the ferrite bead that feeds the -Input to the DOA card. So now all of the channels are working great.

One thing I am contemplating looking at is the pan law in this console. As I said before, it is about 4.5dB more when panned hard left or right, and this is added to if you send to a bus that is also panned. For now I'm going to leave it alone and figure out how to work around it.

Here's a pic of it all together

 

[Potato Cakes]

And are the sexy screws. I didn't fasten down the bus channels yet as I am waiting for the EQ boards to show up and do some testing.

Thanks!

Paul

 

[Potato Cakes]

I finally got the 553 boards for the bus EQ. Everything fit and worked the first time. There's a couple of things I might tweak but it sounds great and I like it quite a bit over the stock buss EQ. The frequency curves are near identical to the CAPI 553F that I used for comparison. I don't have an original but I think Jeff's version for anything API sounds fantastic so I figured if I'd be fine using that as a bench mark. I had to make a parallel capacitor array to get the value needed for the LF (39nF). I noticed that I'm lacking about 2-3dB of gain for the LF section compared to the 553F and I don't know if the capacitor array being used is doing anything unexpected, although the sum of the three caps being used is within the tolerance range (41nF). But it's still getting to right at 12dB of gain, which is more than enough for a majority of applications.

I think I'll swap out a couple of bus channels at a time instead of all at once. Or maybe groups of 4 as I am using the nicer inductors from the CAPI site and they are about $26 a piece, which adds up quickly when you need 32 of them.

EDIT: The LF is sorted. I now am getting  the same level as the 553F. I had to make some adjustments for the cap and resistor following the 1H inductor.

 

[Potato Cakes]

Some frequency response comparisons. The signal routing for the console was channel - buss - stereo. There are garAM16 opamps for the input section and the ACA-Bo board, so the console has a cumulative roll off happening (I think) going through three op amps. This shows up more in the phase measurement what is shown in the response and what is actually audible. The CAPI 553F is direct in/out, using just the one DOA plus the proprietary discrete amp boards Jeff started using in this EQ. Both 553 circuits are using gar2520s.

Blue = TAC stock bus EQ
Orange = TAC 553 mod EQ
Green = CAPI 553F

Here's the HF section.

 

[Potato Cakes]

Here is the MF section.

 

[Potato Cakes]

And the LF.

 

[Potato Cakes]

Also of note is that jennsenmann's IC op amps suggestions were very helpful. The TLE2072's are much nicer in the HF section. I think later I will go back through and change out the chips for the MF2 and LF bands, though they seem to sound pretty good just doing a quick listen. I may not be as satisfied with the sound after I get a chance to do some more critical testing but for now the TL072A's in that position are fine.

With all of the changes, this console is very quiet. The only thing that I'm not happy about is the metering, but I think that would require a brand new circuit to get it to be useable as even the ones that are working don't seem to track very well. It may have always been like that as I know an engineer that used work on this console said he would use the meters on the tape machine as the ones on the console weren't great. For now, the DAW metering will work.

So this concludes my journey, as far as information gathering goes. It honestly wasn't as difficult of a project as it could have been. I think the hardest and most time consuming part of the process was making decisions on what to actually do. Once I had a pretty clear idea of which changes to make each step seemed to go pretty quickly. Some point I'll post a summation of everything that was done in case to aid others undertaking this project to expedite the rebuild process. I thank all who lent me their brains to help me avoid making massive errors and unnecessary alterations.

Thanks!

Paul

 

[jensenmann]

Well done, mate! Congratulations !!!

 

[Potato Cakes]

Thanks! I'm glad this one is done for the most part.

I went ahead and ordered TLE2072A's for the other bands of the channel EQs. I figured since they smooth out the top end there's no reason to not do so for the rest of the bands.

I also noticed when doing more testing that signal going Tape In - Bus - Stereo Out I get unity gain with the fader set to the top of the throw (as I had planned), but Channel - Bus - Stereo Out yields approximately 0.6-0.9VAC more that unity. I'm not exactly sure why, but the rotary faders on the buss channels allow me to dial in the desired level exactly, so I figure I'll just use those when I need to use the bus channels in stereo pairs. It's a little puzzling but I don't know if I'll spend any time investigating.

Also did some initial listening tests with the modded bus EQ. It's really, really nice. Since I am using the CAPI ACA-Bo I'm getting the benefit of having a 2503 style transformer on the output, although mine is the high nickel version. I don't seem to miss the EQ itself not having an output transformer because of that.

Thanks!

Paul

 

[Matt Syson]

For what it is worth (late in the day) the SSM2015 in the schematic used on the JP305 (?) sub board has a little quirk that when clipped produces a short burst of HF oscillation while it is trying to be clipped. This was resolved on AMEK desks using the JP305 by changing to SSM2016 which requires either a different value gain pot or a whole new string of gain set resistors on the desks that used switches.
The replacement board was a 'DH' (I forget the number now) but apart from the gain setting was a drop in replacement.TL072 or TLE2072 doth not an insert send make!
Matt S