EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Janaaoi on May 16, 2013, 11:51:33 am
-
Hello,
I want to run an audio cassette deck from a precision linear bench psu for forensic purposes, but don't know how to connect one up. Can anyone help me, and walk me through the necessary steps?
I attach the power supply schematic, and can post some jpgs as required.
-
What's the real goal? Simulate different AC voltages (like what happens if there's a power brownout)? I assume your bench psu is DC. A Variac would be cheaper than a decent AC power source. You can get a 300VA one for <$100.
Run the transport at variable speed? Tascam 112's already have a knob for that.
-
What's the real goal? Simulate different AC voltages (like what happens if there's a power brownout)? I assume your bench psu is DC. A Variac would be cheaper than a decent AC power source. You can get a 300VA one for <$100.
Run the transport at variable speed? Tascam 112's already have a knob for that.
The goal is to emulate one of the main specifications of the JBR Technology's Standard Cassette Deck, that is built upon the Tascam 122 MkIII, in a Tascam 112 MkII. The JBR removes the internal transformer and replaces it with a benchtop psu in order to remove the line 50/60Hz magnetic field away from the playback head and electronics.
I have some difficult tapes that I need to analyse, and sought a better machine with external transformer. The JBR is currently out of production, and very expensive at $6K US, but I managed to find a new-old-stock 112 on eBay - hence a plan to acquire and modify it.
I don't yet have a bench psu, but will seek something high end, linear, low ripple etc. I don't know quite whether I need research quality - possibly not. I just need a very stable noise-free output that will not interfere with my tape frequencies.
I attach a copy of the JBR brochure, so that you can see their specification.
-
Why not just place the transformer in a remote case and connect it by a lead set and a socket/plug on the unit. Use a steel or aluminium box for the transformer and ground the case and you have removed most of the magnetic fields. Use 470nF on each of the secondary leads with a ferrite bead, with the one end of each capacitor connected to the case, along with a mains input filter and transient suppressor will handle a lot of mains conducted noise into the unit as well.
You need a 10 pin connector, and cable, but that is quite easy to do, use a 15 pin D connector and shell, it can easily be fitted into the case where the original power switch and socket was. Will handle the power required, and you just have to buy some 10 core shielded cable to connect them, however long you need.
-
Why not just get a decent deck that runs on DC, like a Sony TC-D5 or WM-D6C?
-
Why not just place the transformer in a remote case and connect it by a lead set and a socket/plug on the unit. Use a steel or aluminium box for the transformer and ground the case and you have removed most of the magnetic fields. Use 470nF on each of the secondary leads with a ferrite bead, with the one end of each capacitor connected to the case, along with a mains input filter and transient suppressor will handle a lot of mains conducted noise into the unit as well.
You need a 10 pin connector, and cable, but that is quite easy to do, use a 15 pin D connector and shell, it can easily be fitted into the case where the original power switch and socket was. Will handle the power required, and you just have to buy some 10 core shielded cable to connect them, however long you need.
I attach a picture of the transformer. There are no visible secondary leads above the board, so I don't know quite how you would extend them?
-
4 JIS screws and unsolder the transformer, then mount it in a remote case and connect the pins as appropriate to the existing pads. Power cord to the remote case only and disregard the mains switch on the cassette deck, place one on the remote case. It is a pretty well shielded transformer though, you probably would be better off using a mains isolating transformer on the input power line to get rid of mains noise.
-
4 JIS screws and unsolder the transformer, then mount it in a remote case and connect the pins as appropriate to the existing pads. Power cord to the remote case only and disregard the mains switch on the cassette deck, place one on the remote case. It is a pretty well shielded transformer though, you probably would be better off using a mains isolating transformer on the input power line to get rid of mains noise.
I understand, but it is a lot of effort if I could simply ignore the onboard transformer and inject the appropriate voltages externally. Plus I can buy a better quality external transformer that is already in a box.
I used a Qube isolation transformer previously but I have to say it was completely ineffectual. There was no observable difference in any spectrogram produced.
The accepted wisdom is to use an external power supply of some form, that is the only way to achieve really stable audio signals for analysis. Note the JBR machine was designed by James Reames, ex Director of the FBI audio forensic laboratory after thirty years as an investigator. The JBR machine probably represents an ultimate specification.
-
Hello Janaaoi,
I noticed that the Signal to Noise ratios of the JBR and Tascam cassette decks are very close.
Both of the S/N specs are for Metal tape at a 3% THD reference point with all Dolby noise reduction turned off...
The JBR literature shows 60dB
While the Tascam 112MkII spec sheet states >59dB
The transformer in the Tascam deck does look fairly well shielded. Are you sure that you really need to modify the Tascam?
Also, it seems that the Wow & Flutter specs are similiar too...
JBR: 0.06% weighted rms @3KHz (at standard speed)
Tascam: <0.04% wtd rms @3KHz
Just a thought.
Regards,
Frank
-
Hello Janaaoi,
I noticed that the Signal to Noise ratios of the JBR and Tascam cassette decks are very close.
Both of the S/N specs are for Metal tape at a 3% THD reference point with all Dolby noise reduction turned off...
The JBR literature shows 60dB
While the Tascam 112MkII spec sheet states >59dB
The transformer in the Tascam deck does look fairly well shielded. Are you sure that you really need to modify the Tascam?
Also, it seems that the Wow & Flutter specs are similiar too...
JBR: 0.06% weighted rms @3KHz (at standard speed)
Tascam: <0.04% wtd rms (test freq. not stated)
Just a thought.
Regards,
Frank
Hello Frank,
Yes, the JBR is just a modified Tascam 122 MkIII, whilst the Tascam 112 Mk II is a less specified version of the 122. Most of it relating to recording, which is not required in forensic audio, where everything is about playback.
The only real difference of the 122 to me, is the slightly better tape transfer mechanism that it possesses - that is direct drive without any rubber band, and it came as standard with balanced audio outputs. Both of these Tascam can still supply if I need them. So effectively I have the basis of a new 122, that is no longer available, in my 112, thanks to eBay.
My first task to copy the JBR forensic specification, is to feed it with an external power supply - hence the post.
-
Janaaoi I can see what JBR is trying to accomplish. You want there to be as little residual inductive line frequency and other noise contributed by the playback deck as possible. You need to move all of the primary and secondary AC lines out of the deck and feed only filtered DC into the deck chassis. I can see where to split the power supply for the +12 volt and -26 volt supplies, which would be right after the rectifiers and main filter capacitors, and before the regulators, but need to find out what the "F" AC lines feed, since the circuitry there should also be moved off-deck. Can you show the schematic area for the CONT M board? I'm guessing this will have the logic/microprocessor power supply of 5 volts on it. I feel you could use the existing power transformer. As said before, you would need to unsolder it from the CB and mount it upside down on a chassis and solder wires to the pins. You will also need to make up a board to hold the rectifiers and filter capacitors. Multipin connectors and AC input circuitry will finish the job.
One other thing I've noticed on the 112 and 122 decks is the spindle drive motor can induce some low frequency commutation noise in play mode. This was improved in the latest versions of the decks by better motor shielding.
Interesting project!
-
Janaaoi I can see what JBR is trying to accomplish. You want there to be as little residual inductive line frequency and other noise contributed by the playback deck as possible. You need to move all of the primary and secondary AC lines out of the deck and feed only filtered DC into the deck chassis. I can see where to split the power supply for the +12 volt and -26 volt supplies, which would be right after the rectifiers and main filter capacitors, and before the regulators, but need to find out what the "F" AC lines feed, since the circuitry there should also be moved off-deck. Can you show the schematic area for the CONT M board? I'm guessing this will have the logic/microprocessor power supply of 5 volts on it. I feel you could use the existing power transformer. As said before, you would need to unsolder it from the CB and mount it upside down on a chassis and solder wires to the pins. You will also need to make up a board to hold the rectifiers and filter capacitors. Multipin connectors and AC input circuitry will finish the job.
One other thing I've noticed on the 112 and 122 decks is the spindle drive motor can induce some low frequency commutation noise in play mode. This was improved in the latest versions of the decks by better motor shielding.
Interesting project!
That is exactly what I am trying to achieve. However, I am not an electrical engineer, so I am having to feel my way and learn as I proceed.
I attach a couple of photos of the CONT M board. If they are not adequate, or I have misunderstood what you sought, please advise and I will try again.
I identified D21, D23 and D25 as possibly the relevant diodes. D23 is the four wire black rectifier to the right of a large capacitor and the on/off switch.
-
I attach a copy of the CONT M PCB ASSY as a reduced size jpg.
-
Janaaoi,
I noticed that your Tascam seems to have a single 4.8cm/sec speed (1 7/8"/sec), with a +-12% pitch/speed adjustment range.
The JBR version has 4 speeds (1.2, 2.4, 4.8, and 9.6 cm/sec), plus the ability to continuously vary the speed from 1cm/sec to 10cm/sec in 1/100cm/sec steps (very nice).
Will you need to add this sort of very flexible playback speed adjustment feature to your Tascam to have it meet the requirements of your work? if so, that would add an interesting twist to your project. There may be additional playback EQ curves needed for each of the 4 playback speeds.
Frank
-
Janaaoi,
I noticed that your Tascam seems to have a single 4.8cm/sec speed (1 7/8"/sec), with a +-12% pitch/speed adjustment range.
The JBR version has 4 speeds (1.2, 2.4, 4.8, and 9.6 cm/sec), plus the ability to continuously vary the speed from 1cm/sec to 10cm/sec in 1/100cm/sec steps (very nice).
Will you need to add this sort of very flexible playback speed adjustment feature to your Tascam to have it meet the requirements of your work? if so, that would add an interesting twist to your project. There may be additional playback EQ curves needed for each of the 4 playback speeds.
Frank
It would be nice to have the playback speed option, which is another of the great specifications of the JBR machine, but it is sufficient for now to use software to make any speed adjustment to files once on the pc.
The 122/112 decks do have the pitch control of +/- 12% for small adjustments, but you will observe the speeds you quoted are multiples of each other that are easily reproduced with software.
I should add, adding the option myself is a bit beyond my capabilities at present.
-
I have looked at the information you posted, Janaaoi, and you will have a challenge to fully separate the power supply. Are you confident in your desoldering and circuit building skills?
I can see now that the "F" lines are the filament voltage for the vacuum FL display. It has a -26 volt voltage applied through a center tap to cause the filament to act as a cathode, which is typical for this type of display. The potential problem with this is the need to allow that AC voltage to enter the unit. It is only around two or three volts, so that may be acceptable. The big improvement is getting the power transformer out.
From what I can tell so far, you need to remove and rebuild on a new board/chassis: The power transformer and it's associated AC mains circuitry, - D20 and D21 (part of the power up/down mute circuit), - D23 and C21, - D25, R30, C24 , D22.
There is an AC secondary line running over to the MAIN PCB, which probably feeds the split power supply for the audio section. There will be a rectifier or two and probably two capacitors there that need to be moved to the new board as well. If I saw the schematic for the MAIN PCB power section, I could identify those. As part of good design practice, you should add small value electrolytics and film bypass capacitors at the point the external wiring meets the deck circuit boards to maintain stability. These could fit into the positions left by the removed capacitors.
This is no trivial project. It would be great if you could enlist the help of a knowledgeable local person to help you out with this, as it it difficult to convey all the info through a forum!
-
I have looked at the information you posted, Janaaoi, and you will have a challenge to fully separate the power supply. Are you confident in your desoldering and circuit building skills?
I can see now that the "F" lines are the filament voltage for the vacuum FL display. It has a -26 volt voltage applied through a center tap to cause the filament to act as a cathode, which is typical for this type of display. The potential problem with this is the need to allow that AC voltage to enter the unit. It is only around two or three volts, so that may be acceptable. The big improvement is getting the power transformer out.
From what I can tell so far, you need to remove and rebuild on a new board/chassis: The power transformer and it's associated AC mains circuitry, - D20 and D21 (part of the power up/down mute circuit), - D23 and C21, - D25, R30, C24 , D22.
There is an AC secondary line running over to the MAIN PCB, which probably feeds the split power supply for the audio section. There will be a rectifier or two and probably two capacitors there that need to be moved to the new board as well. If I saw the schematic for the MAIN PCB power section, I could identify those. As part of good design practice, you should add small value electrolytics and film bypass capacitors at the point the external wiring meets the deck circuit boards to maintain stability. These could fit into the positions left by the removed capacitors.
This is no trivial project. It would be great if you could enlist the help of a knowledgeable local person to help you out with this, as it it difficult to convey all the info through a forum!
Oh dear! This plan is beginning to falter!
I guess I have become used to plug-and-play devices. Just plug in an external psu. "How difficult could it be?" - I thought!
Nothing is impossible with time and inclination I suppose. Time for a rethink.
-
Here is the main pcb schematic in two parts.
-
And here is the main pcb as a reduced jpg.
-
Okay. Had a look at the schematic of the MAIN PCB, and there is indeed a rectifier bridge, D7, and two electrolytic filter capacitors, C27 and C28 that should be moved out of the deck to your external board. These are for the positive and negative unregulated rails for the audio circuitry. If you can do this, plus what was mentioned in the previous post, you would get virtually all the AC line noise inducing components out of the deck. The idea is to split the power supply between the unregulated, but filtered DC and the voltage regulation circuitry. It will require a multi conductor cable and connectors. I have reused sturdy DB-9 and DB-15 cables for this kind of purpose, since they are cheap or free, at least around here.
If you are able to copy the wiring on the schematic to lay out your outboard circuit board, you should be well on your way. This could all be done on perf-board, I expect.
-
No need to move rectifiers or caps - you can feed them with DC instead of AC.
-
Here is a photo of the 112's boards and internals in reality.