Toshiba SD-3990: audio out bypass -- help needed with mods

[Nerull]

They're used to mute any kind of noise that might appear on the outputs at start-up. Pretty trivial really, but hey, I guess it could annoy some people.

~Tom

 

[headfone]

Okay, after a little bit of squinting and working with oversized test leads on my Fluke 87, here's "rough-draft" -- but, TTBOMK, electrically-correct -- trace:




Larger, higher-rez image here.

Important question:

Does the Zoran (analog 2-channel audio DAC) output current or voltage? If the latter (voltage-output), then that opens up the possibility for mods like Swenson, etc. (I searched for a Zoran chip datasheet but could not find one).

Another question:

What are A in- and B in- pins on the op-amp used for?

 

[Nerull]

Actually, you shouldn't hope for voltage-out because that means that it was I/V (current to voltage converted) inside the chip, which is not good by audiophile means (they usual use cheap, crappy methods of I/V). This also limits how much of a benefit you can possibly get from modding.

I would try calling up Zoran to see what the outputs are, because I can't find anything on the internet about that chip at all (bunch of vague crap at their website).

The A- / A+ and B- / B+ are the differential inputs to the op-amp.
You can find out the hows and whys of it here:
http://williamson-labs.com/480_opam.htm

Or here if you like in-depth analysis

http://focus.ti.com/lit/an/slod006b/slod006b.pdf

~Tom

 

[headfone]

Quote:

Originally Posted by Nerull Actually, you shouldn't hope for voltage-out because that means that it was I/V (current to voltage converted) inside the chip, which is not good by audiophile means (they usual use cheap, crappy methods of I/V). This also limits how much of a benefit you can possibly get from modding. I would try calling up Zoran to see what the outputs are, because I can't find anything on the internet about that chip at all (bunch of vague crap at their website). The A- / A+ and B- / B+ are the differential inputs to the op-amp. You can find out the hows and whys of it here: http://williamson-labs.com/480_opam.htm Or here if you like in-depth analysis http://focus.ti.com/lit/an/slod006b/slod006b.pdf

Great educational resources -- thx!

Because I'm pretty much a rookie at this stuff, I assumed the voltage-output preference was "necessary" as it was suggested in this article for Toshiba players. I assume that's what Swenson also had in mind.

Will the mod in IEEE article above (aka the Swenson mod) work in either case -- i.e. current or voltage out? Also, instead of going the purist -- no opamp -- route, how about using BUF634 in place of the 4558? I think referenceaudiomods.com is using this approach for their Toshiba's.

I'll report back if Zoran shares something interesting.

 

[headfone]

Quote:

Originally Posted by headfone I'll report back if Zoran shares something interesting.

I contacted Zoran by both phone and email. Customer service sucks -- they don't have a formal tech support and one must email all queries to the same general address: sales@zoram.com Sigh. What a way to run a railroad! Anyway, email communications with Zoran are partially detailed below (email Mr. Young at Zoran to voice your concern!!)...

Dear [headfone]:

[first reply to my request]

Regarding your request Zoran does not provide data sheets other that to
manufacturers of end products. We suggest contacting Toshiba for the
information.

[as I noted earlier to Mr. Young and in my first message in this thread, Toshiba does not have a service manual for this model. Contacting Toshiba on this issue, they told me to go to ... you guessed it: Zoran]

[second reply to my request]

I am sorry this [datasheet request] will not be possible for the reasons I specified in my earlier email.

Regards,
David W. Young

Sr. Manager
Sales and Business Development
Zoran Corporation
1390 Kifer Road
Sunnyvale, CA 94086
email:david.young@zoran.com
Office: 408 523-4241
Mobile: 650-248-5346

 

[ECM]

I performed some mods on my sony DVD player and the easiest way for me to bypass the muting transistors was to remove the coupling caps only.

The output from the opamp goes to the coupling cap, then the transistors, then the RCA jack. I removed this coupling cap and installed ONE end of a Solen cap in the hole that connects to the opamp output. Then I snipped the rear of the RCA jack to break the connection to the circuit board. Using a fine copper wire, I soldered the other end of the Solen cap to the rear of the RCA jack and was done. No need to desolder or mess with all the surface mounted stuff.

So the output of the opamp went thru the solen cap straight to the RCA jack.

I also changed the opamp from 4558 to an OPA2227 and biased it into class A with JFETS.

Sounds a lot better now.

 

[headfone]

Quote:

Originally Posted by ECM I performed some mods on my sony DVD player and the easiest way for me to bypass the muting transistors was to remove the coupling caps only. The output from the opamp goes to the coupling cap, then the transistors, then the RCA jack. I removed this coupling cap and installed ONE end of a Solen cap in the hole that connects to the opamp output. Then I snipped the rear of the RCA jack to break the connection to the circuit board. Using a fine copper wire, I soldered the other end of the Solen cap to the rear of the RCA jack and was done. No need to desolder or mess with all the surface mounted stuff. So the output of the opamp went thru the solen cap straight to the RCA jack. I also changed the opamp from 4558 to an OPA2227 and biased it into class A with JFETS. Sounds a lot better now.

Thx for your feedback.

I'm still a bit in the dark about the resistors and caps *between* the DAC and the opamp in the *stock* Toshiba (see image of circuit here: http://www.gammaelectronics.com/Tosh...990.html#trace). I *think* the resistors here may be considered "load resistors" (here's a popular DIY DAC that uses load resistors in the same spot as the stock Toshiba: http://www.dddac.de/ma_dac22.htm). But that still leaves the caps that are perpendicular to these resistors.

The IEEE Spectrum article for the Toshiba mod makes no mention of these in-between components, though. How about in the case of your Sony? Do you have a "straight-wire" path between the DAC and OPA227? And, also, why did you opt for the OPA227 as opposed to other potentials (e.g. BUF634, OPA237, etc.)?

A million q's -- I know! Just want my mod to be right the first time thru

Thx again.

 

[Nerull]

The capacitors perpendicular to the resistors are for filtering out high frequency noise from the DAC. As such, I wouldn't remove them (there would be no audiological benefit anyway).

~Tom

 

[ECM]

Since I wasn't sure what output the DAC in the Sony used, I left everything between it and the opamp as is. I didn't want to experiment with I/V conversion or anything of that sort since at the time, it was my only source for playing CD's.

Using the pinout of the opamp, which is standard 8 pin, I traced the output to the first component which was the electrolytic. After that, I installed the Solen cap and went straight to the RCA. I didn't mess with the muting transistors or load resistors.

I used the OPA2227 because that was the only sip chip i had on hand and it turned out to sound pretty good. I did a mod on an Apex (wal-mart) DVD player and used an OPA2134 in it with good results.

To bias the opamp, I used two jfets and traced the negative power supply trace to an open pad and used it for -V.

Take a look at this post with pictures of the mods I did:
http://www6.head-fi.org/forums/showthread.php?t=119323

With the power supply mods, it also improved the video quality of the unit. My son has watched "Nacho Libre" on it everyday this week!

 

[headfone]

Quote:

Originally Posted by Nerull The capacitors perpendicular to the resistors are for filtering out high frequency noise from the DAC. As such, I wouldn't remove them (there would be no audiological benefit anyway).

Okay -- now I'm confused (probab. from my lack of understanding the Swenson mod and/or the 3950 topology on which that mod was performed (recall, I have the 3990)).

In the Swenson mod, one is asked to take a few caps "surounding the DAC" out of the circuit. Not sure if these are the same perpendicular (high-freq. filtering) caps you are suggesting to leave in. If so, then the Swenson mod is *not* possible -- correct?

Thx again for responding.

 

[looser101]

I hate to dash your hopes for a Swenson mod, but to me it looks like the DAC in the Zoran has balanced outputs. You've traced 2 rows of resistors back to the Zoran but there are an additional 2 rows which I suspect also connects to the it.

If you look at this datasheet on page 34 you will see that the output stage looks very similar to the one on your board.

-R-

 

[headfone]

Quote:

Originally Posted by looser101 I hate to dash your hopes for a Swenson mod, but to me it looks like the DAC in the Zoran has balanced outputs. You've traced 2 rows of resistors back to the Zoran but there are an additional 2 rows which I suspect also connects to the [DAC]. If you look at this datasheet on page 34 you will see that the output stage looks very similar to the one on your board.

Great catch -- I was wondering what those extra/"parallel-running" set of resistors were. I traced the opamp's Ain- and Bin- pins back to the DAC and, yup, they do indeed connect to a DAC adjacent to Ain+ and Bin+, respectively. And this also further clarifies Tom's message.

I'm not sure if having balanced output from the DAC is a "good thing" -- so perhaps someone can comment on that. IAC, I'll probab. go the same path ECM chose with his Sony.

ECM: dunno if you caught this from earlier in this thread ... but I *removed* the output transistors quite some time ago. Sonics-wise, their removal didn't make that much difference.

 

[ECM]

I missed the part where you removed the transistors.

The reason I didn't remove the ones in the Sony was that I didn't want the signal going thru the trace even though the transistor was removed. It just made more sense to me to have it go straight to the jack after the cap. Either way, in the case of the Sony, it's pretty trivial since it's a low end DVD player.

And now you know why I avoided the DAC altogether! The digital stuff gets tricky fast.

 

[PeAK]

Quote:

Originally Posted by headfone Okay -- now I'm confused (probab. from my lack of understanding the Swenson mod and/or the 3950 topology on which that mod was performed (recall, I have the 3990))... If so, then the Swenson mod is *not* possible -- correct? Thx again for responding.

Headfone....I've summarized some of the recent posts/links/threads on the SD-3990 and got inspired enough to pop into my Big Box Retailer and picked one up $65 Canadian].

I'm an EE who performed a modifed Swenson Mod with a scope at home so I should be able to sort out the guts of the SD-3990 and post to my web page...should be up by next week factoring the time to weed out infant mortality and getting a handle on its stock sound. So be patient.

I'm doing this one for you guys/gals as I'm quite happy with my Auricapped SD-3950 a la Swenson...but it be cool if the rumours of it beating modifed units in stock form where true...about a week!!!


cheers,
PeAK

 

[headfone]

Quote:

Originally Posted by PeAK Headfone....I've summarized some of the recent posts/links/threads on the SD-3990 and got inspired enough to pop into my Big Box Retailer and picked one up $65 Canadian]. I'm an EE who performed a modifed Swenson Mod with a scope at home so I should be able to sort out the guts of the SD-3990 and post to my web page...should be up by next week factoring the time to weed out infant mortality and getting a handle on its stock sound. So be patient. I'm doing this one for you guys/gals as I'm quite happy with my Auricapped SD-3950 a la Swenson...but it be cool if the rumours of it beating modifed units in stock form where true...about a week!!!

Looking fwd to your 'scope-based analysis.

I wasn't that impressed with the Auricaps tweak. Mechanical damping was worth the time/effort, however.

I've got a couple of parts coming in from Mouser/DigiKey/Newark for this unit, but won't have time to install them for at least two week (earliest). Many other projects on the table have to take precedence first!