Battery voltage and pinout for Tektronix TDS 3000 series oscilloscope

[Recyclojunk64]

The power-supply on my second-hand Tektronix TDS3012 (A series) DSO has blown, and rather than mucking about trying to repair it, I would like to power it with an external DC supply.

The scope already has what appears to be an input jack for an external powersupply, but for some reason it appears to be shorted, so I would like to power it through the battery connections. However, seeing as the scope didn't actually come with a battery (or even a back cover, as it was mounted in a panel) I have no idea of what voltage it needs, or even the pin-out of the battery.

A google search for the pinout and voltage of the battery proved fruitless, so if anyone knows could they please let me know?

[cs.dk]

Hi,
According to the servicemanual, the batterypack (TDS3BAT) is a 12V 4400 mah battery. The charger (TDS3CHG) is a 18V device.

The servicemanual is here - Page 114 and forward. http://www.physics.wisc.edu/undergrads/courses/fall2012/623/lab_equipment/TDS3000_oscilloscope%20manuals/TDS3000_service_manual.pdf

Hope it is useful.

[Recyclojunk64]

Thanks, only just got around to fixing it (been using the analogue scope until now). Manual was quite helpful, but didn't give the pin-out, so if anyone else needs to know it's +12v on the pin closest to the power switch, and GND on the pin furtherest from the switch, the rest seem redundant. It's now happily running off a salvaged 12v 4A power adapter.

BTW, would anyone know what sort of floppy drive the TDS3000 series scopes use? Mine is missing, perhaps if anyone knows where I could find the pinout; maybe a laptop floppy drive could be installed? Or alternatively if anyone has removed the floppy drive from their scope for whatever reason and doesn't want it, I'd be willing to buy it.

[MarkL]

I took a peek inside a TDS3054, and the floppy is a TEAC FD-04HF 2300-U.  Also on the label is "P/N 19308023-01".

A quick search shows that it's a 1.44MB slim-line laptop floppy.  It's connected with a single 26 pin flat-flex cable (no separate power cable) and is probably a standard pinout:

http://www.allpinouts.org/index.php/3.5_inch_micro_Floppy_Disk_Drive_26_pin

Good luck in your search.  They seem to be pretty rare but you might be able to find a substitute.

[Tuomas]

Hi, I found the topic while googling for TDS3000 series power supply issues and thought I'd post my experiences here as well.

I recently acquired a TDS3014 that also had a broken power supply. Worked perfectly on battery power.

I opened it up yesterday and it didn't take long to figure out the fault. On the mains side of the circuit there were two 1/2watt 390ohm resistors in series. One of them had failed open. I replaced them with eight 390ohm 1/4watt resistors (four sets of two parallel resistors in series). Power supply works now.

I have attached a picture of the replacement resistors. After taking the picture I added some Kapton tape in there to make sure it won't short-circuit against the surroundings.

[Remseck]

Hi Tuomas

Your hint repaired my power supply, but not my DSO TDS 3014. Short on the Main Board now. Shit!!!!. Maybe some correction to your statement about the value of the blowing resistor. On my AC board it was 390 KOhm, not 390 Ohm. Measured and same with colour code After replacing, the output was around 15V.

Juergen

[Dave Wise]

I know it's an old topic but if I started a new one it would have the same title.

Pinout:

<locating key>
MINUS
NC
DATA
PLUS-TO-CHARGER
BAT-PLUS

The DATA line uses the Unitrode/Benchmarq/TI "DQ" 333 bit per second protocol.
I believe the TDS3BATB (maybe TDS3BATC too) contain a BQ2050 gas gauge chip.  TDS3BAT would be an older BQ20xx meant for NiCd.

When the scope is running from mains and charging the battery, once every second it reads six registers, 0x3 (Nominal Available Capacity High byte), 0x17 (NAC Low byte), 0x3 again (in case it rolled over I guess), 0x1 (FLGS1), 0x2 (TMP) and 0xB (VSB - battery voltage).  NAC increments approximately every five seconds.  The temp field of TMP reads a constant 7 (30-40C) while the gas gauge field incremented once in about a half hour.  FLGS1 returns 0x94 indicating Charging and Capacity Inaccurate.  I don't like CI but maybe the scope doesn't care.

By the way I measured this using the TDS itself   Just wrapped a wire around the DQ pin to bring it out to probe.  The gauge's comm timing is unstable, it's best to use B Trigger in Event Count mode.  Commands are 9 negative edges, returned data is 8 edges.  So you can scroll through the commands and replies by setting the event count.

[Dave Wise]

Cross-posted from "Tektronix TDS3BAT rebuild":

About my TDS3BATB that reads Capacity Inaccurate: after a complete discharge and charge, now it reads Capacity Inaccurate + Valid Discharge.  I expect CI (and VDQ) to go away after one more full discharge, meaning the gas gauge is calibrated.

If you want fast charge but don't have a TDS3CHG/TDS3ION, it looks like a normal bench supply will do.  I set mine to 18V with 1.8A current limit, based on the TDS3CHG spec sheet.  Connecting it to the TDS3BATB external charge coaxial connector, I see about 1.3A which tapers after a couple hours then shuts down to a few tens of mA at charge complete.  I will probably charge them this way - outdoors - if Tek continues to refuse my packs for recall.  And store them in a metal box.  That should be safe.
To extend battery life I will store them only about half-charged.

On initial installation in a scope running off the mains, I see the usual six register reads, then this, once:

read 0x7 (PPD Program Pin Pull Down) == 09 i.e. Prog1 and Prog4 power-on straps were low
read 0x8 (PPU Program Pin Pull Up) == 80 i.e. no power-on straps were high, so Prog2/3/5/6 were all in 'Z' state
(By the way, one thing this means is the cells are coke anode, not graphite anode.)
write 0x8c (VTS End Of Discharge Threshold Select) = 0xA2 i.e. 10.63V
read 0x0c (VTS) to confirm it is 0xA2
(This value was probably stored in RTC RAM and depends on what the scope read on VSB just before it died.)

Disappointed to find next morning that after charge it still reads 0x14 on status.  Maybe another cycle is needed.  We'll see.

[Dave Wise]

On a previous cycle I saw VDQ in the status flags, and I expected to clear CI, but it didn't happen.  I don't know why.

But now I'm not getting VDQ any more, and I think I partly understand why.
The battery is cutting off scope power before the voltage sags down to the End-Of-Discharge threshold.
Or maybe it is sagging so fast at that point that I die between one sample and the next.
This is bad.  The scope controls that threshold, but it keeps on writing 0xA2 i.e. 10.63V even though the battery dies before reaching it.
I may be screwed, because I don't know how to change what the scope writes.
I suppose I can try a factory reset, but I don't want to blow the calibration, so that's a last resort.

I wonder if I have a bad cell.  I'll try again to open the case, so I can access individual cell voltages.
The BQ2050 and suspected companion protection chip BQ2053 do NOT balance the cells.  If there is a balancer, it is something other than those two.
(I don't know what the charger is.  I thought maybe BQ2054 but that chip is buck only and I think we need buck/boost.)

I may end up cobbling a "discharge and DQ comms jig" so I can read the registers the scope doesn't.  And maybe write some.

[Dave Wise]

Got it open!  In the center one-third of the pack there's a foam spacer glued TIGHT to the batteries and the lid.
I sliced it apart with a sharp knife.  Be careful not to hit wires etc!

TDS3BATB CONTENTS

Conspicuously missing: CELL BALANCING!

12x Panasonic 18650HG Li-Ion cells, 4s3p

Candor VC4003A board (date code 23/05): gas gauge, protection,
and switching charger powered by TDS3CHG/TDS3ION.

Candor VC4003B board (date code 41-05): linear charger powered by scope.

VC4003A has wires to BAT-, Cell1, Cell2, Cell3, Cell4, Thermistor,
TDS3CHG +15V and return, and SYS GND, LOAD+, and DQ to VC4003B.

VC4003B passes DQ through to scope connector.

PARTS ON VC4003A:

BQ2050 gas gauge
MAX1737 switching charger
MAX932 dual comparator
Mitsumi MM1414 marked "1414C 529" 4-cell over/undervoltage protection
4x Siliconix 4435 P-channel mosfet SO-8, seem to be connecting LOAD+ to battery
2x Siliconix 4410 N-channel mosfet, seem to be the charger

PARTS ON VC4003B:

MCAF S21B tsop8
LM393 dual comparator
BQ2057W linear charger tssop8
2x Fairchild NDT452AP P-channel mosfet SOT-223 in parallel
Siliconix 4435 P-channel mosfet, switches LOAD+ to scope
Linear charger is powered by flyback boost regulator made of
MCAFS21B chip and 2N03 SOT23-6 ZXM62N03E6 N-channel mosfet.

WIRING

From batt to VC4003A:
BATT- is black
CELL1 is violet
CELL2 is yellow
CELL3 is orange
CELL4 is red
Thermistor is black/black

From TDS3CHG to VC4003A:
+15V is green
+15 Return is brown

From VC4003A to VC4003B:
SYS GND is black
LOAD+ is red
DQ is yellow

Since TDS3BATB does not balance, I conclude my problem is imbalance.
And yep, here are the no-load voltages two days after full charge:
4.11, 4.11, 4.11, 4.21.  Looks like segment 4 is subnormal capacity
so it gets topped off before the others.  CONFIRMED: Segment 4 starts
out high but falls behind the other segments as they discharge, becoming
the first (by a long ways) to hit the Undervoltage trip point.  That is, we
disconnect due to cell undervoltage instead of the normal disconnect
due to pack undervoltage.  That's why the gas gauge can't calibrate.

I may try adding a fourth cell just to find out whether the bad cell is just weak or entirely gone.
Looks like the bad cell is weak not gone, because adding a cell results in that segment having more capacity than the others.

I wonder if this is the root cause of the recall.  Cranking a cell up to its overvoltage cutout on every single charge will eventually blow it up, don't you think?

[Dave Wise]

TDS3BATB interior.  Unlike TDS3BATC, nine of the twelve cells are packed into the bulge, with thermistor on top.

[Dave Wise]

TDS3BATB VC4003A board: switching charger, gas gauge, protection.

[Dave Wise]

TDS3BATB VC4003B board: linear charger powered by scope's mains power supply.

[Dave Wise]

PSA: Anyone still using a TDS3BATB (as opposed to sending it to Tek for exchange) should pry it open and install a balancer.

I wouldn't be surprised if TDS3BAT also lacks balancing.  Maybe that's why MarkL in "TDS3BAT Rebuild" had to keep rebuilding his pack.

[MarkL]

PSA: Anyone still using a TDS3BATB (as opposed to sending it to Tek for exchange) should pry it open and install a balancer.

I wouldn't be surprised if TDS3BAT also lacks balancing.  Maybe that's why MarkL in "TDS3BAT Rebuild" had to keep rebuilding his pack.

My TDS3BAT indeed did not have any balancing circuitry, but then again it was a NiCad pack.  NiCads are much more tolerant to overcharging and typically don't need series balancing.

Not to say a balancer wouldn't have done a better job keeping the pack at a decent capacity.  I'm beginning to think maybe the charging circuit in my pack had a problem given others' success with rebuilding.

And to confirm, the TDS3BAT uses the BQ2014.  There's also a PIC16C54A on there too, two LP339 quad comparators, an LP324 quad opamp, and a bunch others.  There's a lot going on.

[Dave Wise]

The real place where a balancer helps NiCads is on discharge, to prevent cell reversal in primitive packs without voltage monitoring.  In TDS3BAT, the undervoltage cutout should operate before that happens.

[Dave Wise]

NOTE: My TDS3BATB contained Panasonic CGR18650HG cells (orange case) which are rated 1800mAh.

UPDATE: Tek finally replaced my battery packs.  TDS3BATC reads out with the same pull-up/pull-down strapping, which means the gas gauge is less accurate since the new stack is 6450mAh instead of 5400mAh.  (And it was always kind of off due to faulty self-discharge curve because the straps say coke anode not graphite anode.)

I didn't "take it apart", but I did peek inside.  The lid comes off easily, Tek switched to non-adhesive foam.  The cells are Sanyo UR18650A, rated 2150mAh.  There are three circuit boards instead of two.

Board #1, the fast charger, gas gauge, and cell monitor/protection, is 99% the same layout as TDS3BATB.  (A couple of caps are deleted, and one FET is a different package.)

Board #2, the slow charger, is a total redesign.  Instead of a booster and linear charger, it's a switching charger based on the LT1512.

The new board #3 is mechanically and electrically between #1 and #2.  The only wires are red/black pack-plus and pack-minus, and the DQ gas gauge signal.  There's not much on it but I can't wrestle it out to see, everything's RTV'ed in place.  I see a DFN-8, some SOT-23's, and a narrow 8-pin.  I think it's nothing but another layer of SC/OC/OV/UV protection.

JUST LIKE TDS3BAT AND TDS3BATB, TDS3BATC DOES NOT BALANCE THE CELLS.