Batboard

Just a little tidbit for everyone out there.

While switching over a customer from a conventional UHF system to a trunked UHF system this past week, we installed 16 MCS2000 base stations with the HPN4002 Motorola (Astron) switching power supply, as that's what they provide when you order the 'base station' option.

We programmed in the customer's existing repeater frequencies so they could continue to use their existing system until the trunked system was up & running. Within hours of installing the MCS radios, complaints of cross-talk began.

After installing the trunked system yesterday (MTR2000 repeaters), I tested it with a couple of base stations. Not only would the chosen trunked voice channel show a receive light, but so would at least 2 other repeaters, regardless of which channel was chosen! Portables & vehicle installed MCS radios did not exhibit this problem.

I switched the new power supply with an older Astron SL11 we had removed from service a few days earlier...the problem went away.

I haven't had a chance to check it out with a spectrum analyzer yet, but it must look like the NY city skyline when using the HPN4002 power supply...

What really burns me is a local FTR admits to seeing this exact MCS2000 / Astron problem years ago on another UHF system...yet neither Motorola nor Astron has ever bothered to remedy the issue.

We now have to switch out all the power supplies tomorrow. It would have saved a lot of work if Motorola could have at least offered a different power supply for UHF MCS base stations.

I don't know if this applies to VHF & 800MHz units or not.

Todd

I'm not surprsied at all, since /\/\ has practically outsourced all of the experienced engineering personnel from that product group and since the power supply is built by Astron (outsource OEM), the problem will continue.
So much for /\/\'s Digital Six Sigma Quality Program. The /\/\ management believe everything can be corrected in software but not in this case.

Damm....

Thought they learned their lesson with the Astron transformer supplies in the GR series repeaters blowing up. They switched to Duracom 18 amp surge switchers with the fan. Worked great. I ended up changing out quite a few of these things. Wish I could remenber the P/N.. HPN 4033 or something??

Dave

Wasn't there also an issue with exploding Quantar power supplies?

Checked it out with a spectrum analyzer today...yep, there were the spurrious emissions plain as day. In particular, two quite large spikes on either side of the carrier, seemingly exactly 100KHz apart. There was also another smaller spike on either side (for a total of 6, if you're counting).

What a pain in the @ss to replace the new supplies with the old ones we had just removed. Now, if our salesman gets some other type of power supply from Motorola as a substitute, we have to do it again. Anyone willing to bet Motorola is willing to foot the bill for the triple work load? Ha Ha Ha...

Todd

This is exactly why I'm a firm believer in the old linear copper-wound, over weight, over sized transformers for long-term and relaible operation.

Switchers are nice for some applications, but not all.

If you get decent switchmode supplies without spikes everywhere, they're generally great

Personally, I only run linear supplies on my own gear (except one Simoco radio, which lives in a base housing + power supply from the manufacturer).

Weird thing is, we've used hundreds of the things on other models of UHF radios, as well as hundreds of 800MHz MCS2000s with nary a problem. Only seems to affect UHF MCS2000 units for some reason.

Todd

Sounds like they did a poor filtering job on the power supply. Check ripple (A/C) voltage under load. I'll bet it's through the roof. As for the problem with that specific radio, maybe it is more succeptible to ripple than others?

I have a bad feeling the 2 spikes you see 50 kHz either side of the carrier frequency are actually TX sideband noise. Since you haven't specified what kind of a Spectrum Analyzer you're using ... H-P or Tektronix with a good dymanic range of -120 dBm and a good RF Notch Filter tuned to the TX RF channel, it's anyone's guess what you are really viewing.

It could be nothing more than front end overload on the Spectrum Analyzer's receiver. If they are true Spurious signals, you will need to go into the RSS under the RF power adjustment parameters, Key up the TX and vary the Power Output up and down while observing the Spectrum Analyzer screen. The true spurious transmitted signals will walk around in frequency or not staying in the same place as the power is varied.

I'm wondering how DC Amperes are being drawn from the Astron power supplies on the transmitted signals. You could be drawing too much current and the DC supplies are folding back causing them to appear defective.

I really doubt if spectrum analyzer dynamic range is an issue here. There's nothing wrong with looking at a transmitter carrier, passed through a power attenuator [as in a service monitor type instrument], or sampled with a coupler, so long as the user understands the procedure.

Spectrum anlayzer dynamic range is defined as the maximum ratio of two signals simultaneously present at the input which can be measured to a specified accuracy [IEEE 1979]

It is limited by second and third order distortion products in the instrument and by the instrument's noise floor - Displayed Average Noise Level [DANL]. DANL is determined by the analyzers narrowest resolution bandwidth filter.

Once the instrument's optimum mixer level is exceeded [typically -30 dBm], second and third order products, either harmonics of the signal of interest or products generated between the various signals present at the input, will appear to be real signals. To avoid this problem, the analyzer's microprocessor starts adding input attenuation as you attempt to set the instrument's reference level above the optimum mixer level.

For affordable instuments the maximum display dynamic range will be on the order of 80 dB. This is sufficient to determine transmitter spurious emissions as described by wavetar. In this type of measurement, the dynamic range is noise limited as the products being examined are close in - therefore the second and third harmonic distortion in the instrument is outside of the frequency range of interest.

Since signal to noise is the limiting factor in close in measurements, some instruments [e.g. Tek 2780 series, Agilent ESA] will allow you to 'overdrive' the mixer - i.e. increase the maximum mixer level.

The real problem users get into with spectrum analyzer dynamic range is having multiple high level signals simultaneously present at the input - often signals that are outside the range currently being displayed cause trouble. All sorts of signals may be displayed, some real and some internal products.

Internal mixing products can be ruled out by increasing the instrument's attenuator setting. The noise level will rise with input attenuation, real signal amplitudes will remain constant, internal IM products will drop in amplitude or disappear.

The issue here is: how is the crud from the switching supply finding its way onto the transmitter carrier. For example, it could be poor filtering in the power supplies, leaving high frequency components riding on the DC and thus AM modulating the radio PA.

Motorola should know better by now.

Way back when - I had an experience of this type. I had the opportunity to Beta test the then new Pulsar II head and the new UHF Pulsar drawer unit. [OK, OK so WAY WAY back when]

I had them installed in my Corvette. The phone only worked for a mile or so from the main site before the call got noisy. We knew there was desense for some reason but test equipment available in the field was primitive in those days [before the R2001 and before the CE-15]

We had mounted the antenna on the rear of the car opposite the AM/FM antenna, with foil tape on the underside of the body to form a ground plane [Yes - I drilled a hole in a new Corvette]

Motorola said nobody else Beta testing these phones had any problem whatsoever. They decided it had to be the car. As it turned out, Marty Cooper [father of the cell phone] owned a Corvette, so they talked him into installing one and he had the same problem.

They took his car into the RF lab anechoic chamber and put it on the turntable. They determined that the fiberglass body let the RF pass right through and that the mass of the engine and transmission attracted the RF [counterpoise?] to go right that direction. That caused it to pass over the junction box and induce RF into the circuitry which contained a 15 Volt switching supply for the microprocessor in the control head. The switcher operated in the kilo-Hertz range and the mix caused a range of spurs to be generated far enough out to desense the phone receiver..

They completely re-designed the junction box and power supply filtering before the Pulsar II went into production.

All I can say is with a different power supply, the 6 100KHz-spaced spikes go away, and the radio will not try to key up 3-4 repeaters at a time. I don't really care what I'm seeing, as long as it can be rectified one way or the other. I was using an R2670, which I know doesn't match up to an HP, but is certainly good enough to show me the problem.

Todd

Quick update:

Finally got a GCC (global customer care) number assigned for this problem, after the local Motorola branch couldn't/wouldn't help us. Now I have a fellow in Plantation, Florida looking into the problem. He couldn't replicate it with his shop radios & supplies, so I had to ship down a couple of the problem units. He reported today he received the radio & power supply & can replicate the problem. Now I wait to see the fix...

Todd