HELP! SurfPI Constant tone - but hacked to work 'backwards'?

[kt315]

from my BOM file on bara (again bara)

You need be aware that there are two versions of 7660 - the 7660 and the 7660A.
7660 performs supply voltage conversions from positive to negative for
input range of +1.5V to +10V, and 7660A has an input range of +1.5V to +12V.
The maximum supply voltages are also:
7660 = +10V
7660A = +12V
7660C = +10V
7660S = +12V

[fo0bar]

might help or not, but when I made my first surfpi kit from silverdog I didn't have a component tester and I had place the red cap values in the wrong spot on the board it worked but was not working well once I correctly placed them all it worked perfectly. A clear photo of the board could help others with more experience see the problem better.

Thanks Tibuck19 - The components are all certainly in the correct place, and I've definitely done all the checks, multiple times over the board for cold joints/shorts etc, also reflowed many with the SMD blower. - though 'seeing is believing' of course! ...And another set of eyes will very likely catch anything I've missed - so tomorrow I'll be posting a couple of photos when the sun's up. I'm aware that one resistor value is off by a couple of K, but being at the audio end it's noncritical regarding this issue and merely affects the volume ever so slightly.

[fo0bar]

from my BOM file on bara (again bara)

You need be aware that there are two versions of 7660 - the 7660 and the 7660A.
7660 performs supply voltage conversions from positive to negative for
input range of +1.5V to +10V, and 7660A has an input range of +1.5V to +12V.
The maximum supply voltages are also:
7660 = +10V
7660A = +12V
7660C = +10V
7660S = +12V

In this case, the 7660 is being supplied +5v from the 78L05 to generate the -5v - I would assume both models would work for this application, but correct me if I'm wrong!

[Qiaozhi]

In this case, the 7660 is being supplied +5v from the 78L05 to generate the -5v - I would assume both models would work for this application, but correct me if I'm wrong!

In the Surf-PI circuit, the 7660 is being powered from the output of a 5V linear regulator, so either a 7660 or a 7660A is ok. This is also why the -5V supply is lower than -5V.

It would be useful if you could post some scope images of the waveforms at U1 pin3, and at TP2.

[kt315]

your right George.

With R11 and the 100K pot in place I cannot pull the offset down to 0 - on a scope or multimeter. Removing the pot completely has negligible effect. Turning the offset pot seems to have near zero effect, and I can only get an extremely feint resonse from the NE5534. It sits at around 600mv, with very feint, small spikes where the pulses should be appearing.

---
you may check offset without a signal on input. simply disconnect/desolder mosfet and look output voltage.
tune offset by pot. problem again? - write about.
second. do not use big coil for testing in your room. use spool with ferrite core.
probably you have big EMI influence from modern devices on the coil so decreasing the coil size
you decrease the influence in same degree.

[6666]

Even with bad EMI and the coil connected you still should be able to adjust the DC offset + - zero volts, if you got your 5534's from RS they should be real not fake, but you never know, the off set circuit is 2 components a pot and a resistor, try breadboarding the preamp on a solderless plugin breadboard and test a 5534 , just use 4 x aa's for + and - 5v supplys , with all the component replacements you have done something smells fishy.

[waltr]

Even with bad EMI and the coil connected you still should be able to adjust the DC offset + - zero volts, if you got your 5534's from RS they should be real not fake, but you never know, the off set circuit is 2 components a pot and a resistor, try breadboarding the preamp on a solderless plugin breadboard and test a 5534 , just use 4 x aa's for + and - 5v supplys , with all the component replacements you have done something smells fishy.

Good advice and should prove what might be wrong.
Is possible the Pot is bad. You can test it with your Ohmmeter.

[fo0bar]

It would be useful if you could post some scope images of the waveforms at U1 pin3, and at TP2.

As I've removed 'external' components for ease of taking photos, I'll connect it all back up and get some photos of the scope too.

you may check offset without a signal on input. simply disconnect/desolder mosfet and look output voltage.
tune offset by pot. problem again? - write about.
second. do not use big coil for testing in your room. use spool with ferrite core.
probably you have big EMI influence from modern devices on the coil so decreasing the coil size
you decrease the influence in same degree.

I have also tried this by removing the FET and coil. This made no difference to the output/'zero' voltage of the opamp, except that I was no longer seeing the feint spikes where the pulses would have been. While troubleshooting, I've left the FET sitting high on the PCB for ease of removal in case I need to perform this test again.

Noise and interference was my initial concern with this issue, and so I also tested in different rooms of the house, in the garage, as well as the back yard away from any residential or underground powerlines/EMI sources. I have been working on a PI design of my own which hasn't been seriously affected by these factors while in my workshop, but thought I'd eliminate that possibility anyway

Even with bad EMI and the coil connected you still should be able to adjust the DC offset + - zero volts, if you got your 5534's from RS they should be real not fake, but you never know, the off set circuit is 2 components a pot and a resistor, try breadboarding the preamp on a solderless plugin breadboard and test a 5534 , just use 4 x aa's for + and - 5v supplys , with all the component replacements you have done something smells fishy.

I got these NE5534's from eBay, and am now waiting for a few to arrive from RS. I was actually going to breadboard up a test circuit using a 7805, 7905, NE5534 and associated semi's this afternoon, so I'll let you know how it goes.

One thing that might interest you (and I haven't added this previously for the sake of not muddying the situation) is that for *some reason* the 7660 died when I ran the unit with no coil. This made absolutely NO sense to me. I replaced the 7660, tested with coil attached, all ok... removed the coil again and the 7660 burned out again. This got me looking at the -5v as being suspect.

However..... very strangely, when the 7660 was dead and I tested the opamp's output, it was suddenly amplifying the input signal, however I still had no control over the offset. This led me to believe my NE5534's work fine on +5v but once I pull the GND down to -5v and thus operate it on ~10v, I suddenly lose the output again. Unfortunaately with the lack of -5v the rest of the circuit doesn't respond and I get no audio whatsoever, not even a constant tone.

Good advice and should prove what might be wrong.
Is possible the Pot is bad. You can test it with your Ohmmeter.

The trimpot I was initially using wasn't the best - it read 80k between the outer pins and 150R at minimum, so I've been testing with a full-sized pot that reads a perfect 100k/0R.

Anyway, here's the photos of the board - again, please note I have removed the external pots, r6 isn't present as I'm using the variable resistor until I get it all functioning 'more correctly'. The FET sits high for ease of re-removal during troubleshooting and will be corrected later. I'm aware the resistor at the audio end is off by a few K and don't really care about the volume difference caused by it.





I do also have a slightly closer photo of the opamp area if required.
I'll post again shortly with scope pictures once I've got the externals connected again.

Thanks again everyone for your input and help - you're awesome!

[Qiaozhi]

This might be a silly question, but I assume you did check the supply voltages directly on the pins of U2 (pins 7 and 4), and not just at the supply?

I cannot see any mistakes by looking at the photo, so you're next course of action should be to breadboard the preamp and confirm that it's possible to adjust the null offset. Then you can compare voltages between the two circuits and look for that Eureka moment.

[6666]

This might be a silly question, but I assume you did check the supply voltages directly on the pins of U2 (pins 7 and 4), and not just at the supply?

yep I was guuna ask

[waltr]

It is really sounding like those 5534's are Fake and do not have the proper Balancing circuits in side them.

Please post when you get the 5534's from RS.

A side note: many PI detector circuits that use the 5534 do not use the Balance pins. Instead the offset adjustment is a simple summing of a Voltage divider pot into the non-inverting input. This also allows using many other op-amps since many op-amps do not have the exact same Balancing pins. Even the 318 (seen in older PI designs) use the balancing pins differently. This is also way I think the 5534's you got are fakes.

[fo0bar]

This might be a silly question, but I assume you did check the supply voltages directly on the pins of U2 (pins 7 and 4), and not just at the supply?

I cannot see any mistakes by looking at the photo, so you're next course of action should be to breadboard the preamp and confirm that it's possible to adjust the null offset. Then you can compare voltages between the two circuits and look for that Eureka moment.

I did check the voltages directly at U2 - I pretty much traced everything around the opamp to verify all voltages were as expected..

Eureka moment.

....And that moment has arrived


I breadboarded the opamp this arvo; I'm running it on 12v battery through a 7806 which is reading 5.86v at the opamp vin. 22k resistor from +6v to the pot wiper as per SurfPI & datasheet.


The opamp output is reading 359mv at one end of the pot, and 361mv at the other end. This is just over half of the error value of around 600mv I was reading when powering the 5534 at 10v. I don't think this is just a coincidence... It looks like it's definitely a useless batch as they're all acting the same way.



As frustrating as it is to have a bought a bag of stale chips, it's great to know that it's highly likely that no amount of tinkering with the circuit would have helped. Now it's time to patiently wait for the replacements to arrive!

If the replacements do indeed get this thing going, I'm considering throwing one of each in a beaker of acid and comparing the dies under the microscope. While I won't be able to see individual components I can see if the entire layout is different/simplified, which is often the case with counterfeits...

Regardless of this breakthrough, if you do still want to see pictures of the scope at the 555 output and TP2 just to verify, I'll be more than happy to get them for you!

It is really sounding like those 5534's are Fake and do not have the proper Balancing circuits in side them.

Please post when you get the 5534's from RS.

I'll certainly let you know how it goes when the parts arrive, and if I can get photographs of the dies I'll post those too - but in the meantime I do want to say THANK YOU to everyone for all of your input and guidance - you're legendary!

[waltr]

If the replacements do indeed get this thing going, I'm considering throwing one of each in a beaker of acid and comparing the dies under the microscope. While I won't be able to see individual components I can see if the entire layout is different/simplified, which is often the case with counterfeits...

I'll certainly let you know how it goes when the parts arrive, and if I can get photographs of the dies I'll post those too - but in the meantime I do want to say THANK YOU to everyone for all of your input and guidance - you're legendary!

Wow, I'd like to see photos of the dies.

Just for giggles, education and since you are waiting for new parts:
On that breadboard re-wire to SUM the pot into the non-inverting input to adjust offset.
Check other PI detector schematics, like the Hammer Head, etc. for the offset circuit.

[fo0bar]

Wow, I'd like to see photos of the dies.

Just for giggles, education and since you are waiting for new parts:
On that breadboard re-wire to SUM the pot into the non-inverting input to adjust offset.
Check other PI detector schematics, like the Hammer Head, etc. for the offset circuit.

Sounds like fun. I like experiments

Had a look over the hammerhead and replicated the offset circuit along with its relevant resistors (100k/1M)..... 358mv, and moving the pot doesn't offer any mv of difference:



I also tried with 22k resistors, exactly the same result.

[6666]

There is nothing wrong with the MM is there ?