Allen and Heath Saber Modifications

[Greg]

Thanks Gyraf, but just to pick your brain a little, why did you suggest changing that 10u to 470n polyester in the HPF? Which I believe is in the signal when the HPF is off.

 

[gyraf]

That's because the 10uF at that point (because of it's high load impedance) was way larger than needed for audio..

Jakob E.

 

[mshilarious]

Hey all. I ripped apart my new Mixwizard to see what was inside. Looks like a simplified version of the same circuit. I'm gonna do the cap swaps listed above, and swap the opamps for BBs.

One thing I definitely don't like is the fader is too easy to move. I like 'em a little tighter. Anybody have a replacement part to recommend?

 

[Greg]

Just to post some feedback. I finally got a pair of modded channels back in my Saber. Here's what I changed:

- 330uF to 2200uF Panasonic FC
- removed tape and line coupling caps
- swapped eq bypass 'off' cap to 470nF polyester
- bypassed output cap with 100nF polyester

I listened to a variety of 2-track material through them, and the response differences are subtle, but they are there. I noticed a slightly "tighter" and "less muddy" low end. It just seemed to hold together a little bit better down there. Also, on some material, I noticed that it was easier to hear details in guitars in the mid to low-mid range. The top end seemed to be about the same.

There are still more caps that I have my eye on, but I'm in the process of converting another pair of channels, but this time I'm going to use a Nichicon PW Series cap instead of the Panasonic FC I used.

I haven't done any IC swapping as of yet.

After I get this pair done, I'll have enough modded channels to maybe throw up a kick, snare, and 2 overhead drum mix and compare with 4 unmodded channels. Hopefully I'll be able to hear some "not so subtle" differences.

These mods (ie cap swaps) are very easy and inexpensive to do... I'd recommend something similar to just about everyone.

 

[mshilarious]

I just got my Mixwizard back together tonight! So far I've done:

- Opamps on 4 channels (OPA2134 for TL072)

- Cap swaps/bypasses similar to Greg on 4 channels (Nichicon UPW bypassed with poly films)

- Cap swap/bypass on master faders

- Selected opamp swaps (again OPA2134 for TL072) on master channels (aux busses, AB out). However I left all the summing 5532s in place. I get nervous about swapping that out, and it seems to be a decent opamp. I also left the 072s on the stereo returns and headphone amp, since those didn't seem important. I didn't change any other caps on the master channels.

Initial testing of modded channels against stock channels showed an obvious improvement. I'll listen some more tomorrow.

From here, I know I'm going to do the cap swaps on all the other channels, then compare with my opamp swap channels. I'm a bit hesitant about swapping more 072s since OPA2134 is a much hungrier chip. I might be back to ask some advice on the power supply.

 

[BradAvenson]

[quote author="NewYorkDave"]Hmmm... you've got me thinking. In the case of a simple single-opamp inverting stage, you could use a dual opamp configured as one inverting followed by one non-inverting stage, enclosed within one feedback loop. Then you'd have metric shitloads of open-loop gain to give you plenty of feedback margin.[/quote]

Oh and this is a nice page combining that with a lundahl output transformer:
http://www.klaus-boening.de/html/output.html

It seems that at a certain gain it works fine, but when you are changing the gains, (ie adding or removing items from the busses) you may have stability issues. (previously brought up)

 

[mshilarious]

I had a look at the power supply tonight, the regulators & ICs are all rated at 1A or more if heatsinks are used. The transformer I can't find the part, its marking is AM3476. Familiar to anybody?

Since I have the board out, I'll probably heatsink just for some insurance. As for caps, there are 4 10uF 400V caps before the primary side, and a half dozen 470uFs after the secondary. Is it worth upgrading for quality or size? There are a few ceramics I'll change too.

I'd estimate stock is less than 300mA. Given that each OPA2134 adds 2.5 mA of max draw, any guesses how many I could safely swap? The 4 channels I did sound really nice, I'd love to get at least 8 of them (total 24 chips).

 

[longsoughtfor]

I have been getting some distortion through the headphone amp built in to my A&H Saber console. I haven't had a chance to look at the schematic yet (Hi Greg but thought I would pull the board out and take a look anyway.

The realized that both the headphone amp and the main monitor amp were TL072's... bummer. I had some OPA2134's in the cabinet so I popped one in for each amp.

A truly amazing difference: Highs and mids are much clearer and overall, better sounding.

This solved the original headphone problem but my real question is: Is this OK?

The mix bus outputs on this console are NE5532's but the monitoring is through a TL072... why would someone design a console with one amp driving the two bus and an inferior amp driving what you hear? Now I believe I have reversed the difference, the monitor amp is better then the two bus. I guess it is like upgrading to better monitors and I will need to do a couple of mixes to see if they translate better.

As a side note, I monitor through a pair of Event TR8's. While repairing one last week, I saw that the power amp in these speakers is basically a TL072 buffer, a TL072 used as the active crossover and two LM3886's, one for each speaker. I may pull the TL072's out of those in favor of OPA2134's... why not? I'm sure the (unregulated) supply in the speakers can deal with it.

Cheers
Kevin.

 

[Svart]

i've done the same thing, tl082s where there should have been something better, replaced with opa2604 or opa2134, or even dual opa627s. any of them are much better replacements!

i'm thinking of doing a small adaptor from two 2604s to a quad output to replace tl084s and the like..

 

[Greg]

hmmm... I bought a bunch of caps some time ago to recap 4 channels to see what improvements could be made. Unfortunately, I haven't gotten around to actually making the changes. I haven't messed with any opamp replacement either. I think you might need to watch out for the possibility for HF instability... maybe ?!?

Does the OPA draw more current? If so, if you start replacing many opamps you might need to check out the PSU. This is just something that crossed my mind.

 

[Svart]

the OPA stuff does draw a little more current, not much more but it would add up as you add parts.

yeah do check the PS, mine is a switcher running around 60% so I've had a lot of fudge area to add more, better amps.

 

[longsoughtfor]

Does the OPA draw more current? If so, if you start replacing many opamps you might need to check out the PSU. This is just something that crossed my mind.

Yes it does and I'm not suggesting replacing all TL0x parts or any of the 5532's for that matter, but the two 072's I replaced in the monitor circuit do sound better now.

I can't tell from the channel (310) schematic what is IC1?

Kevin.

 

[mshilarious]

[quote author="Greg"]hmmm... I bought a bunch of caps some time ago to recap 4 channels to see what improvements could be made. Unfortunately, I haven't gotten around to actually making the changes. I haven't messed with any opamp replacement either. I think you might need to watch out for the possibility for HF instability... maybe ?!?

Does the OPA draw more current? If so, if you start replacing many opamps you might need to check out the PSU. This is just something that crossed my mind.[/quote]

I now have 8 opamp & cap upgrade channels and 8 cap only upgrade channels. The difference between the two is subtle, I think the opamps add improvement, but I've had trouble controlling all the variables. It isn't the same dramatic difference I got vs. the stock channel.

I wish I'd done opamp only channels for comparison, but I've had this thing open for two months and I needed to get finished . . .

The 2134 draw 2.5 mA peak more than a 072. I've been cautious with my Mixwizard board since upgrades & haven't used all 16 channels. I'm awaiting parts for PSU improvement.

As for instability, 2134 is a pretty conservative opamp, it isn't extremely fast, just a bit more, mainly it has less noise and much less distortion than 072, so I had no troubles at all.

I didn't change the 072 for the headphone since the quality of the headphone jack is not that important to me, and I wanted to conserve power. I'm more concerned with the channel direct outs and aux busses since this is a recording console for me.

In the Mixwizard IC1 is an 072, it wouldn't surprise me if the Saber was the same since the circuits seem similar, even if the boards do not.

 

[Greg]

Moved to the "Drawing Board" ... sorry...

 

[synthdr]

Hi Gyraf I have been working on my allen heath Saber for a while I have recapped the master section I am curious about the 220uf 16v that sit in the L or R mix section of the L and R pcb's of the master section. are there any changes you would suggest like increasing the UF  or am I barking up the wrong tree.
Synthdr

 

[jensenmann]

If these caps sit in the audiopath directly, then increase them and add bypass caps in parallel with 0,1uF like Wima MKS2. This will lower phaseshift in the audioband.