Protek 608: Fried by a "69"

[k1mgy]

30 years in electronics did not insulate me from this beauty - something you should NOT try at home.  I determined that I should use my (once-functioning) Protek 608 handheld DMM to measure its own battery voltage.  Why?  Well, why not??  I simply wanted to know if it was time to replace it.  Selecting VDC and attaching the probes across the 9V cell yielded a sput sound and a blank LCD display.  Dead as the proverbial doornail.  Doing a "69" with a multimeter is apparently not a good move.

It would be lovely to resurrect this meter, but a schematic is no where to be found.  Anyone listening have access to it?

Or perhaps some tips on where to begin looking.  Empty head calling for help.

No burn marks or odours to follow for this one.  It's a mystery.

[Nerull]

It's kind of amazing just how many stories there are of this exact sequence of events.

[electr_peter]

Short answer - do not do it.

Such topic was discussed before on this forum dmm-reading-it's-own-battery-voltage-low-voltage-detection/

[k1mgy]

I've taken it apart (in Dave Jones speak), found no issues, re-assembled and now the meter powers up.  It's been quite a while since the initial damage and my memory of the event may be fogged in so perhaps it powered up before -- or maybe there was a mechanical issue.  Anyhow, it has life, but is otherwise no good GOP.

Symptoms:

1. DC mode: 0.0000v indicated and does not move with input; 10.50 Megohm shown in secondary display and does not change when shorting leads
2. AC mode: 000.00 mv indicated and does not move with input; 1 Gigohm shown in secondary display and does not change when shorting leads
3. Ohms mode: 00.00 Ohms indicated; 2.5V shown on secondary display (1/2 of 5V vcc?); LCD display dims slightly when shorting leads with no other change shown
4. Capacitance mode: main display shows no digits and is constantly toggling between uF and nF; 2.5V shown on secondary display (1/2 of 5V vcc?); LCD display does not dim when shorting leads as it did in Ohms mode
5. Hz mode: display moves about in a range around 30Hz with 000.00 V on secondary display.  The LCD display dims slightly every second, and just before the Hz display is updated.  Holding the POSitive lead causes the display to read around 120Hz
6. Diode mode: 00.000 V indicated; 1ma on secondary display; Shorting leads dims LCD display slightly;
7. Microamp mode: 0000.0 uA indicated; 100 ohms on secondary display; Shorting leads has no effect
8: Milliamp mode: 0000.0 uA indicated; 100 ohms on secondary display; Shorting leads has no effect
9: Amp mode: 00.000 A indicated; 0.01 Ohms on secondary display; Shorting leads has no effect

The meter has a single A/D converter, a Cirrus Logic CS5508BSZ, which is a nice but expensive (around $20 usd) part.  Might be a reasonable gamble to replace it, or perhaps (a bit tricky due to meter construction) lift the input and test the device independently.  Although, given the symptoms being so widespread, it seems reasonable that the damage is in this part.

Anyone care to speculate along with me?  Anyone have access to the schematic/service manual?  How about a cheaper source for the A/D converter?  Looks like Protek is out of business, which is unfortunate as this meter is a mid-range cost, high quality unit with a good set of specs and features.

[Alex Eisenhut]

Is it your only meter? Because the paradox is you should measure power supplies first.

[ModemHead]

... you should measure power supplies first.

Absolutely, the best place to start.



Measure from the COM jack to both battery leads. I don't know anything about this meter specifically, but analog ground is usually regulated to a point in between Vbat- and Vbat+, often about 3V below the Vbat+.  The damage was likely done when the COM lead was connected to the battery terminal.  Might have zapped the power supply circuitry.

[retiredcaps]

I've taken it apart (in Dave Jones speak), found no issues, re-assembled and now the meter powers up.

I'm no expert, but I'm learning about how DMMs work and I can offer suggestions.

Can you post clear focused photos of your pcb (front and back)?

Symptoms:

1. DC mode: 0.0000v indicated and does not move with input; 10.50 Megohm shown in secondary display and does not change when shorting leads
2. AC mode: 000.00 mv indicated and does not move with input; 1 Gigohm shown in secondary display and does not change when shorting leads
3. Ohms mode: 00.00 Ohms indicated; 2.5V shown on secondary display (1/2 of 5V vcc?); LCD display dims slightly when shorting leads with no other change shown
4. Capacitance mode: main display shows no digits and is constantly toggling between uF and nF; 2.5V shown on secondary display (1/2 of 5V vcc?); LCD display does not dim when shorting leads as it did in Ohms mode
5. Hz mode: display moves about in a range around 30Hz with 000.00 V on secondary display.  The LCD display dims slightly every second, and just before the Hz display is updated.  Holding the POSitive lead causes the display to read around 120Hz
6. Diode mode: 00.000 V indicated; 1ma on secondary display; Shorting leads dims LCD display slightly;
7. Microamp mode: 0000.0 uA indicated; 100 ohms on secondary display; Shorting leads has no effect
8: Milliamp mode: 0000.0 uA indicated; 100 ohms on secondary display; Shorting leads has no effect
9: Amp mode: 00.000 A indicated; 0.01 Ohms on secondary display; Shorting leads has no effect

For 1, 2, 7, 8 and 9, the primary and secondary interface are displaying correct, according to the manual, for no input.  Obviously, it should show something other than 0.0000 with input for 1 and 2.

For 3 and 6, the primary display should show 0L, overload, with no probes attached.

From the above observations, it almost seems like something is shorted for 1, 2, 3 and 6 causing it to display 0 all the time.

The meter has a single A/D converter, a Cirrus Logic CS5508BSZ, which is a nice but expensive (around $20 usd) part.

I haven't seen this ADC before so I will have to do some reading/studying.

As a simple test, test VDC with 10V input.  See if the network resistor input divider show the appropriate ratio drop.

In addition, you might want to measure the current draw of the multimeter.  According to the manual, it should be 7mA for most modes.  If the current draw is signficantly higher, then something is shorted.

PS. The battery life for a zinc carbon 9V is 36 hours according to the manual.

[k1mgy]

Excellent suggestions all 'round.

As to voltage measured at the leads of the defunct meter:

DC/AC Volts: 0 v
DC/AC MV: 2.65v
Ohms: 2.24 (moving around between 2.2 and 2.3 v)
Capacitance: Had to examine on scope.  A sweeping triangle wave that cycles every 1 second or so.  Baseline sits at +1.5v.  Triangle wave base is around 0.5v.
Hz: On scope, a funcky looking complex waveform that seems to cycle between square, triangle and sine waves with a center at 0v dc and excursions of +- 100mv or so
Diode test: Sits solidly at +9v DC on test leads
Microamp, milliamp and amp scales: 0.0 V dc

Current Draw: 7MA on average on all modes.  I reported display dimmed when shorting test leads in certain modes.  In these cases, current draw rises to only 8ma.  I expected much greater draw in this condition so perhaps shorting is dimming display through other means.

To access pins on the A/D chip, I attached a few wire wrap type wires to various pins.  None of the attachments were made directly to the device in order to avoid any damage.

I brought out:

AIN +   Analog In   
AIN -    Analog Gnd (or differential)
VREF +  A/D Reference           
VREF -   A/D Reference Return
VREF OUT  Internal A/D reference out
VA +  A/D Reference
VA -  A/D Reference return
DGND  Digital ground
VD +   Digital power

Note: DGND and VREF - are tied together (perhaps at star ground point)
The VREF OUT and VA - are tied together, indicating the internal 2.5V reference is being used in the circuit

Values with no input, DC scale:
DGND to VD+: 3.3V
AIN - to AIN +:  0V
Note: VREF - is tied to DGND
VREF- to VREF+ : +2.5V
VREF - to VREF OUT: +2.5V
VA- to VA+ :  7.15 V

AIN - to AIN + Measurements
0 V input:  0V
10 V input: 2.94V
3V input:  1.3V
1.3V input: 0.567V

Notably, disconnecting the 9V battery (leaving the meter selected to DC Volts) I see 0V between AIN- and AIN+ with 10V across the test leads.

According to the data sheet, "no pin should go more positive than VA+ (+0.3V).  Unless there's a voltage doubler (or otherwise a step-up converter used) then putting the unit's internal battery of +9V at the inputs should not, it seems to me, violate this.  However, "ground" being virtual, perhaps all bets are off? 

Perhaps no concern with the 7.15V seen between VA- and VA+ as between VA - and VREF - I see +3.3V; the same for VA+ and VREF -.

So, what is this telling me? 

1. The flow from test leads to the A/D appear to go through either an analog switch or an active device.  Regardless, the level seems fine (at least at 10V)... however the values are NOT proportional with the inputs over a small range between 0 and 10V DC
2. As there IS a difference voltage between the AIN+ and AIN-, and zero on the meter, I suspect the ADC.
3. The rest of the ADC setup looks OK.
4. I did NOT look at the digital side.  I suppose a quick check to make sure we have clock and some data activity would be good.

Without schematic, no telling what's happening otherwise.  My conclusion is to replace the $20 ADC chip if the digital measurements yield data activity.

[k1mgy]

Is it your only meter? Because the paradox is you should measure power supplies first.

Not the only one.

Now, that WOULD be quite a problem, eh?  Like "physician, heal thyself".

[free_electron]

its dead.

digital multimeters typicall make an artifical zero point to whihc the adc is refrenced. by measuring it's own battery voltage you have shorted this refrence generator to one of the meters own rail ( that doent happen if you measure something 'outside' as one lead goes to the reference point, the other to the input. measuring internally you short the meters own power rail to the ref generator.

so you fried the reference generator of the meter. if you are lucky this is a 2.5 volt zener diode somewhere. if you are unlucky this generator sits in the main chips and that guy is now dead..

happy new year ...

[k1mgy]

Is the reference generator synonymous with the ADC internal analog reference?  This ADC Vref is 2.5V according to my testing.

Unless there is another reference I should look for beyond the ADC?

[retiredcaps]

Looks like Protek is out of business, which is unfortunate as this meter is a mid-range cost, high quality unit with a good set of specs and features.

From this thread

http://www.antiqueradios.com/forums/viewtopic.php?f=8&t=58409

Hung Chang made Protek.  Hung Chang is renamed to GS Instrument.  The 608 is listed on their product webpage.

http://www.gsinstrument.com/02_product/sub01_01.php?board=board01&classcode=B01-4

[Billy03]

The Protek 608 was sold rebranded as Voltcraft VC608 in Europe.
You can still download the schematic on the vendor's website:

http://www.produktinfo.conrad.com/datenblaetter/100000-124999/120277-sp-01-en-Digitalmultimeter_VC_608.pdf

I've fried my 608 too... and had to make heavy use of the schematic.   

[lowimpedance]

Hah does look like a 2.5V ref diode. But what an awful schematic !.

[philtulju]

Sad! I hope you all were able to do something.

Unfortunately the internet is way worse now in 2023 vs. 2014 or 2017 when this thread was active and the schematic link was live.

I did the same boneheaded thing and smoked my Protek 608 that is my favorite meter and that has served me well for more than 20 years.

Fortunately, I was able to fix it - in my case only one component, Q1, popped. Looking at the board Q15 and Q16 appeared to be the same component - SOT-23 marked big capital "N" then "OM" capital but smaller (or possibly "0M") and a date code sideways of "QB" capital.
Looking up the code was fruitless (got some hits for other components like the ZC that seem plausible; Korea Electronics 2N3904S NPN.)
Anyhow I took off Q15 and when measuring it noticed that it was behaving like an N-channel MOSFET. So I replaced my Q1 with a BSS138 that I had in my stockpile and my meter is back from the dead! YAY!

Now I've put a label on all my meters to help remind me:

[!] CAUTION [!]
self-measuring battery
can damage meter