TASCAM M-3500 ajusting voltage rails....with pics and schematics

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

pacemaker

Well-known member
Joined
Jun 8, 2006
Messages
585
Location
France
Hi Guys,
because of lack of money,
i have to continue using my tascam M3500 rather than changing it,
i have allready recapped the whole board and PSU,
changed grounding to star grounding,
changed opamps to 2604 in the preamp section, OPA2132 in the buss sections
and NJM5532 elsewhere,
I now would like to change the frequency of the low freq eq and high freq eq,
wich are 100Hz and 10KHz ,to 80Hz and 12KHz
i also would like to upgrade transistor in the preamp and EQ sections,
the mic pre transistor are NPN 2SC1844
http://pdf1.alldatasheet.com/datasheet-pdf/view/108512/NEC/2SC1844.html
the EQ transistor are 2SC3112
http://www.toshiba.com/taec/components2/Datasheet_Sync//50/6481.pdf
here are the scans of the tascam input :
http://img4.imageshack.us/i/tascaminputa.jpg/
http://img299.imageshack.us/i/tascaminputb.jpg/
any help for these mods  would be greatly appreciated,
especially changing the frequencies and transistor EQ.
Thanks for reading my post,
best regards,
Francois
 
Hi ,
sorry i don't have pics but it's the standard star grounding mod as stated in this article:
http://www.tangible-technology.com/articles/200b.html
regards,
Francois
 
Ok
for the low freq i suspect R141 390K and C124 0.0039uF on the input A pcb
to set the frequency,and is C123 100uF a coupling cap after the pot ?
for the hi freq i suspect R140 1.8K and C122 0.0082uF;
do i need a simple RC filter calculator like this one:
http://www.opamplabs.com/rfc.htm
to find the correct value ?
If i use the calculator with R 390K and C 0.0039uF
it gives me 104Hz so  considering tolerance etc...
it seems ok for 100Hz ,as annouced by tascam.
so for the desired 80Hz it gives C 0.005uF
and with R 1.8K and C 0.0082 , Freq is 10782Hz so roughly the 10K announced
Am i correct ?
If am on the right way,axs i would like to use standard cap value
I would like to know if changing the cap or resistor value is the same ?
Hope someone could help,
best regards,
francois



 
pacemaker said:
I now would like to change the frequency of the low freq eq and high freq eq,
wich are 100Hz and 10KHz ,to 80Hz and 12KHz
You need to change the caps C122 & 124 to 6.8nF and 4.7nF respectively. I'm not sure it would make much difference. How did you come to this choice of frequencies?
i also would like to upgrade transistor in the preamp and EQ sections,
the mic pre transistor are NPN 2SC1844
I don't think you can find anything that will bring any real improvement. The transistors here have been designed at a time where manufacturers had a real market in high-end audio, with moving-coil phono cartridges in sight. These have not been bettered yet. In fact, I think any improvement in semiconductors will never be enough to overcome the basic flaws of the simplistic design. So, unless you're committed to making a serious modification of the mic pre (ccs for the input transistors and a 2-4 opamp complement), you cannot expect any further improvement.
the EQ transistor are 2SC3112
This circuit here is a simulated inductance (sometimes improperly called gyrator). The resulting inductance value is C124.R141.R143 = 2.7H with a DC series resistance equal to R143 and a parasitic parallel resistance equal to R141. The only drawback of this scheme is that the transistor noise is added to the signal. The only way to relly improve this circuit is to replace the fake inductor with a real one, a costly and bulky option, because the inductor must be built on a large size core and shielded against external stray fields. For the same reason as before, changing the transistor is a blind alley.
 
Bonjour abbeyroad,
merci pour ta reponse !!!!!!

This choice of frequency came from my small experience,
i think 100hz is a bit too boomy, and i wanted something lower,
but if you have other suggestions i'm open minded.
i also saw in different forums that other people have done this mod,
and liked it pretty well.
Finally i choosed this values as a starting point for tweaking,
Compare the modified channels with stock ones in A-B situation
or make a small ghetto board with a switch
and compare caps value and material.
I've done this for my pultec build,
and i liked ,as a lot of people,
polyester or polypropylene caps for the low freq,
and styroflex for the high freq.

Thanks for your answer about transistor too,
I found while looking at datasheets and comparing them,
these 2SC1844 and 2SC3112 seems pretty good,
So i will let these in peace and build more outboard mic pres.

If you have a few time and would agree to help,
i'm also looking for a way to change the bus output level from -10dB to +4dB
cause i don't have the LA-3500 balancing kit ;-)
 
If you change C123 to 1.5uF, the low freq eq bandpass will be centered very close to 80Hz and about 1 octave wide. No other change needed.

Try one channel and see if it sounds okay, as the bandpass will much more narrow than the current 100Hz eq band (which is very broad).

regards, Jack
 
mmmmmmmmmmm

Thanks for the tip Kir-for-pleasure !
nteresting boards
i also thought about jeff API summing kit's
just bought a 2-ACA to change the summing amp/balancing boards on the master section,
i allready tweaked a few things like elna caps and opa627 on the summing section and balancing board,
was a nice improvement.
but i expect the API master section to be a bit higher ;-)
I hope to customize this console as far as i can !
Thanks for eq tip AMZ-FX i will try this ,
But from 100uF to 1u5F ? are you sure
?
Best regards,
Francois
francois

 
pacemaker said:
But from 100uF to 1u5F ? are you sure

The current bandpass is a very broad peak at 10Hz (very broad) that lifts up the whole bottom end as the tone knob is increased. When you change to 1.5uF you get a 1 octave band centered on 80Hz.  The capacitor change moves the center frequency up and increases the Q.  If it is not what you like, you can easily put the 100u cap back, or try something in between, like 22uF, which will move the freq some and narrow the Q but less than going to 1.5u -  With 22u, it will still be a fairly broad peak affecting most of the low end.

regards, Jack
 
awesome answer !
i will put some wires,switches and listen to different position to choose what i prefer,
What do you think about testing different caps material for low freq and High freq
polypro for the low and polystyrene for the high ?
best regards
Francois
 
Peacemaker please come back to this thread and write about what you done with the desk..intresting !
i bought a couple of that 1646 for balancing line input but the 1646 only amplify 6db so i must come up with something more for the other 8db...and lacking time for reserch...i dont know which type of opamp connection i should use inverting,noninverting....
 
Hi Guys,
a few help is needed again,
i'm trying to figure how much current the previous JRC4580
were drawing from the PSU,
There was 32x5=160 chips just for the input channels,
I found this datasheet for JRC4580,
http://www.datasheetcatalog.org/datasheet/newjapanradio/ae04056.pdf
annoucing 50ma output current wich seems a lot for me
and the NJM5532 replacing them seems to draw 10mA
Am i correcct thinking i have plenty of juice left ?
( 40mA x 32 = 1,28A ?? )

 
aie !
i had a look at the schematics and datasheet of the different part of the PSU
the +15/-15v rails and +12v rail are powered by a 10D4B41 diode bridge
http://www.datasheetcatalog.org/datasheets2/12/120262_1.pdf
wich is rated between 10A and 4.3A depending on temperature
and rectified by 2SD845 and 2SB755 wich are rated for 12A
http://pdf1.alldatasheet.com/datasheet-pdf/view/123354/TOSHIBA/2SD845.html
wich seems a lot for me ;-)
there's roughly 200 chips in the board
if 4580D draw 6mA that would be 1.2A ,
don't know if it's per rail or total...
but it seems a lot lower than the PSU possibility
if you have a few spare time to help me calculate the power of this PSU
so as i can learn and be sure the PSU won' t blow...
Regards,
francois
 
While the quiescent current of the opamps is significant because they are are always drawing, you also need to add in the loads they are driving under signal, meters, and LED status indications, etc.

Probably not a bad idea to modify one channel as you would like it and measure that as compared to an unmodified channel to get a feel for where you are going. Measurements could be as simple as cutting a 10 ohm resistor in series with +/- supplies at the channel.

Perhaps not a bad idea to beef up the PS. You can get some extra output with more capacitance and heatsink. It may be easier than you think... or not. If your console model came with different channel loads, and yours is not the largest, you may already have extra capacity in your transformer.

JR
 
thanks for your answer Johnn,
the led's drivers meters etc are powered by another power rail
i will scan the PSu schematics so as you can see how it's done,
But this PSU seems very beefy
1 2SD845 for the +15v
1 2SB755 for the -15v
both rails power feed only the opamps
1 2SB755 for the +12v
and 1 2SC3181 for the 48V
I can't find the transformer datasheet unfortunatly
 
here are the scans of the PSU

http://img267.imageshack.us/img267/4968/psu1r.jpg

http://img153.imageshack.us/img153/6541/psufinal.jpg
 
i found this schematic kir for pleasure
you might find it interesting
 

Attachments

  • impedance_balanced_output.JPG
    impedance_balanced_output.JPG
    202.9 KB · Views: 58
  Input A pcb-Change Q101 and 102, to 2SC2546,U101 to Lm6172,C112,C 113 22pf,C114 to 220uf with .01 poly in parellel,C104,C105 to 47uf with .01 poly, U102 to LT1358,R111 to 243ohms,.01 poly across C111,C112 to 22pf.Kudos to Jim Williams for this one.
 

Latest posts

Back
Top