SE2 problems...

All posts related to SE1, SE2 and SE3 and our future stereo encoders go here!

Moderators: Sir Nigel, Nina, pcs, 5r, phpBB2 - Administrators

Post Reply
GrasulGras
New registered user
New registered user
Posts: 5
Joined: Thu Jan 26, 2012 8:19 pm

SE2 problems...

Post by GrasulGras »

Well... i don't know how to start this. I really don't mean to offend or to be disrispectful and i do hope my message will not be deleted..

This is something between a review and a complaint, regarding my newly bought se2000.

I bought this one for my own personal use, not for a radio station but for casting my own music to an available frequency in FM, on a dummy load.
Why ? Because i have a hobby in creating presets for sound processors and i love the way well modulated music sounds on the air, in the 3 meter band.

So, having some technical experience and having built three coders myself, i bought the SE2000 because it promises >60db separation.

The coder arrived quite fast, very well packed and PCS were as serious as i expected them to be.
To be honest, at first look the build quality was a little poor, but i can only claim that because some capacitors on the board were not that "beautifully" aligned.

Strictly an aesthetical problem. As soon as i got home and set up the coder, i adjusted the +6.25 KHz deviation as specified and - voila - it works flawlessly. I mounted the recommended 1 KOhm pot. as indicated in the manual. Playing with it improved separation, indeed, especially > 1 KHz. Here are the results, measured :
not exactly the commercial radio separation i expected, but - nevertheless - good quality. I measured ~ 40dB of separation @ 1 KHz on an in-car Panasonic radio-casette
player, ~20dB separation @ 100 Hz, and ~43 dB at 10 KHz. The results were fair, especially at this price.

The input stage is unbalanced although it has RCA connectors (normally used for balanced inputs). However, it works well if you simply connect your sound processor to
the input.

Used for about 3-4 hours total (with breaks). To my astonishent - it started... well, dying. Slowly. Let me explain... Yesterday, as i powered it up a red led lit up on the board. Seems like a clipping led, not sure what it shows. The manual does not say anything about it. Also, there was no 19 KHz carrier. I thought to myself it could be a charged capacitor somewhere. As soon as i connected both audio inputs (L+R) the 19 KHz carrier appeared and the coder started working very well. BUT, if i disconnect the channel on which that led lit up, the coder spits out a loud signal, unlocks the PLL in the exciter and.... guess what, it's unusable. OK, i thought to myself, unbalanced inputs... high imput impedance, whatever. BUT, the same does not happen if i disconnect the other channel... The coder stays stable and works well.

OK, whatever, i'll deal with it later i thought to myself. This evening, as i powered up, the usual, no 19 KHz pilot as the coder powered up. I connect the sound processor and - both leds light up. The audio sounds horrible and the level is low. Both leds light up, even with the audio feed connected. Sometimes the coder just spits out overmodulated signal, unlocking the PLL, sometimes it spits out a strange, poorly modulated 38 KHz carrier with a (3/2L)-(R) signal. What i get, with exactly the same setup as the first day, a 2 dB diffence between the channels and a monofonic signal. If i move the PCB around the signal either improves either breaks back down. just like a flawed contact.

I can only assume it's a faulty line somewhere on the PCB, but i'm not sure. There's no schematic and this device differs a lot from the "Modulate the 38KHz carrier with a balanced modulator" design... so i'm not really sure how this works (except for the LPF and perhaps the SSM compressor chip at the input). How the 38KHz carrier is modulated - beats me :)

But that's not the point... the point is - that i spend 90 EUR's on something that broke (with or without my fault). I don't mean to criticise PCS, this message will probably get deleted, but i hope not. I just did my best to write an objective review about this coder and hope to get an honest answer. Don't know if it has warranty or not. There's a 30 day return with money back, but i'm not sure if i can return it if i stuck the 1 KOhm pot. in it (it should probalbly be returned as bought). Maybe someone from PCS will shed some light...

Thanks !
GrasulGras
New registered user
New registered user
Posts: 5
Joined: Thu Jan 26, 2012 8:19 pm

Post by GrasulGras »

correction, it's se2000+, not SE2
GrasulGras
New registered user
New registered user
Posts: 5
Joined: Thu Jan 26, 2012 8:19 pm

Post by GrasulGras »

PS:

Here's a recording of how it sounds when you feed both channels, disconnect + reconnect one channel (all audio dissapears) and handle the PCB.

http://www.mediafire.com/?8s8fanqvbzvrahp

(4,7 MiB Wave File)
GrasulGras
New registered user
New registered user
Posts: 5
Joined: Thu Jan 26, 2012 8:19 pm

Fixed It

Post by GrasulGras »

Well, fixed it after some tinkering. For some reason, both grounds of the RCA connectors failed. Faulty PCB i suppose. I soldered a wire from both RCA connectors to ground and it now works flawlessly. Both leds are off, sound as good as the first day.

In case this does not get deleted, i uploaded a file to mediafire with an exact unedited capture of how this coder sounds.
Perhaps. From how it sound like, i think the 19 KHz pilot is a bit too high, as i'm forced to feed a stronger signal to get good modulation. Either the pilot level is too high and i adjusted the exciter to be less sensitive (resulting in an overmodulated 38 KHz subcarrier), either this is all it can deliver. Im really curious to see some response on this one. Screwing around with the "Pilot Level" pot seems to have no results on the pilot level itself :(

Here's an actual recording made with a sony NWZ-B3 walkman.
Mind the fact that the clicks and crackes in the background are from the recorder ! The stereo coder's output is very clean.

http://www.mediafire.com/?p6g032duzu86org
User avatar
pcs
Site master
Site master
Posts: 3132
Joined: Fri Jan 18, 2002 1:00 am
Location: Radio Land
Contact:

Post by pcs »

Hi,

we don't delete messages unless they contain advertising material, faul language or similar so no worries there. Let me address your questions:

1. The inputs are DC coupled balanced inputs, this is to ensure perfect low frequency response and noise suppression. There is one drawback to such inputs, when the ground of your audio source and ground of your transmitter are not at the same DC potential, you will have a problem. For example when you have non-grounded mains powered equipment the filter/circuitry in the mains power supply can produce DC offset towards ground and this DC offset can vary with time. When you touch such equipment with your hand you will sometimes feel a very gentle shock or if you go over the enclosure with your hand just barely touching it you will feel a weird "electric" feeling, sorry don't know how to explain it otherwise. It is basically some low-current mains voltage. Medical equipment has higher standards so "medical" specced mains power supplies have far less of this problem as a sensitive patient could have a problem.

Simple way to handle this is to either:
a) connect both grounds with a wire
b) use grounded 3-prong mains cable in both units (always recommended for protection anyway).
What you did by grounding the RCA connector is basically a). This does reduce asymmetric noise suppression though which is non-issue for you probably.

2.) The pilot level trimmer does in fact have some impact.

3.) If you think the quality of assembly is poor I have to strongly disagree there, you must have very high standards. Here's an actual photo (not photoshopped):
http://www.pcs-electronics.com/images/p ... 00v2_b.jpg
I advice you to open some products in your household, you may be surprised what you will find there :)

4.) If you are looking for top stereo separation check this out, its a brand new model with completely new multiplexer:
http://www.pcs-electronics.com/se6000-s ... -2008.html

To my knowledge this is highest oversampling rate I've seen in such coders.
Best regards,
Marko - PCS Electronics
--------------------------------------------------------
Turn your PC into a FM radio station!
http://www.pcs-electronics.com
fax +386 4 2316 128
GrasulGras
New registered user
New registered user
Posts: 5
Joined: Thu Jan 26, 2012 8:19 pm

Post by GrasulGras »

Dear Marko,

thanks for the answer and tips. I'm sorry about being a pain in the ass, i did not mean to offend.

Thanks for the tips.
It would be nice to have more documentation sent with the coder itself and some sort of block schematic
(i do realise your schematics are trade secret) but as long as you have a MCU inside, the schematic
itself is sometimes worthless to reproduce without the MCU code. Instead, it could be helpful for
solving small problems like these, eventually improving some issues (19 KHz pilot varied only slightly until
i replaced a resistor with a wire, on the board. Then i had more control over the pilot and re-adjusted it.

But, this is only a thought.

One question stands, though. Agreed, the potential difference between the two grounds (x-mitter and processor)
could be (or definitely is) the culprit. In this case, why did the coder have problems with the processor
disconnected (both inputs blank, one (eventually two) led's light up. I suppose they could be clipping indicators...
The 19 KHz carrier would shut down completely. I did not connect it to a scope to see what happens exactly
with the output, still, the fact that the "STEREO" led on the receiver did not lit up was enough for that
moment, as the PLL remained locked.

As for assembly, yes, i have very high standards :) I like all the parts to be mounted at perfect angles,
this is purely aesthetical. I've seen a lot worse, but, i have also seen better. Better can only be achieved by robots in mass
produced stuff, but i have seen PCB's assembled by hand that looked dazzling.
They were not from a competitor of yours :D

Anyway, it works good enough and you can consider me a "satisfied customer" :)
User avatar
pcs
Site master
Site master
Posts: 3132
Joined: Fri Jan 18, 2002 1:00 am
Location: Radio Land
Contact:

Post by pcs »

Hi,

ok, let me see if I can answer your questions:

1. The SMD parts are laid out by a "robot", however since whole board is one big ground (this is recommended for RF design) one end of the SMD component sometimes stands in a large sea of "mass". It is due this large mass of copper on one end that sometimes in the owen they can be pulled slightly to the side. So as the components are placed they are all aligned perfectly, once you bake them they could be pulled to the side slightly. You will not notice this problem in a design without large areas of mass/ground.

2. We limited pilot adjustment range to prevent users getting into problems (most don't have equipment to measure level).

3. Once you applied high DC level offset to the inputs of the opamp you probably pushed inputs of some opamp into latchup. They have zeners on inputs and once activated they can act a bit like thyristor, they short until you power off and back on. It can be accompanied by whole thing drawing a lot of current. I think you are lucky the inputs survived.

M
Best regards,
Marko - PCS Electronics
--------------------------------------------------------
Turn your PC into a FM radio station!
http://www.pcs-electronics.com
fax +386 4 2316 128
Post Reply