Orban and innovonics audio kit...

[SamTheDog]

Hi all.
For a while I've had an Innovonics 230 series broadcast processor. This has been used here with the redifon BT500
and a smaller mediumwave transmitter. Now I've picked up an optimod 9200AM. this was made redundant from a bbc
site I believe. I've been testing this with a low power mediumwave transmitter here, and wondered if anyone else has
had experience with these? It seems to be holding the peak-level in the mid to high 90% range. but pushing it up the final
few percent decreases the "clarity" of the broadcasted signal...

Any tips and tweeks greatefully received !

PS.. the optimod has v1 software.. Anyone got a later eprom image they can send me?

Cheers
S-T-D

[Albert H]

You have to remember how the Orban kit works - it's just glorified clippers! The "clarity" really suffers when you get much beyond the onset of the clippers. Bob Orban tried (very hard) to minimise the distortion problems by using a multiband approach, but ALL clipping is distortion, so you're really just adding a slightly improved Fuzz Box to the end of your airchain!

At Inovonics, we made a conscious decision to avoid clipping as far as possible. The limiters generally used PWM control, as this makes a wide range of control possible, with great channel-to-channel tracking (if you're doing stereo), and an absolute minimum of clipping artefacts. That's a big part of why the Inovonics gear became the Industry Standard (the other major issue was price!). Your 230 uses FET-based differential attenuators (which were a compromise, of course). I never really liked them, since they would introduce second harmonic distortion at quite mild gain reduction levels. Many people liked the "concordant" distortion, but it was anathema to my sensibilities! However, the distortion never really exceeds 1% (the Orban can be >10% in some circumstances!).

If you look at the circuit, you'll see that the channel filters are all transistor-based (rather than op-amp). This was done for reasons of current consumption reduction (and price)! The shape of the filters is actually pretty good, despite their simplicity, and the overall baseband - when they're all recombined - is remarkably flat.

I never liked the "Phase Follower" function of the 230 - it seemed an unnecessary gimmick - but AM stations cited it as a major reason for purchase. The eight-band compression always seemed too involved, but the competitors were all going for "octave-based" processing, so we had to "go with the flow". Practically, for AM, three bands is plenty (to my ears) and phase rotation trickery can get you another dB (if you're lucky).

I would consider the Orban 9200 to be a downgrade! I've never managed to set one up to sound as good as the Inovonics. Even the (relatively) cheap 222 sounds better.

Have you been through the 230 alignment procedure? It's quite a long-winded operation, but the improvements that can be made by tweaking up an ageing unit can be worthwhile.

I'm going to stick with my first-run 222. It does the job I want, and sounds great on the air. I've done the "European" mods to the filters, and it's still (to my ears) just about the best combination of processing for AM, and if you can find a secondhand one, they're quite competitively priced.

[SamTheDog]

Hi Albert H.
Yes... I imported the Inovonics 231. and I knew it was in need of repair/service. It needed some FETs replacing and still has a few lazy/iffy meters. But apart from that is all working as it should. I went through the alignment procedure in the manual twice ! and it actually sounds pretty good! I picked the orban up quite cheaply from a chap who had been recovering kit and spares from closed-dowm mediumwave sites. Over the past four weeks or so I've been experimenting with it. BUT when I try to get the average mod level from mid/low 90% up to the mid/high 90% the audio gets far worse. I expected some distortion. Let's face it anything in the audio-chain introduces it's own bit... But going from an acceptable amount (to my old ears) to, too much is a narrow margin on the orban.

ps this orban came from LBC according to some of the setups on it..

Cheers S-T-D

[SamTheDog]

What are the European mods Albert ?

[Albert H]

The European mods reduce the audio bandwidth to suit 9kHz channels. There are jumpers on the board for this, but there are a couple of capacitor changes that you can do that improve it further.

You've discovered the Orban secret - it's just "tuned" distortion!

Have you ever played with Aphex? I designed my own version of the Aphex processor, and it's well worth doing. The Aphex processor uses a state-variable filter to select a small(ish) band of frequencies (usually in the 800Hz - 2k5Hz range) and deliberately introduces second harmonic distortion. A small percentage added fools your ears into hearing the selected band as "louder" - it's a great way of making vocals stand out in an otherwise bland mix. Some stations used Aphex to make their presenters sound "louder". It's like the Optimod - if it's overdone, it sounds horrible! If you want a circuit to do the Aphex trick, I'll post a couple up here.

I'll still put my 222 up against pretty much any other MW processor! It was a well-considered design, and the bare minimum to get really great results. They're pretty much the Industry Standard in the USA.

[SamTheDog]

I've never used APHEX Feel free to enlighten me..... !

also, on a semi-related note...
Have you ever used a dolby A module modified to a "vocal stressor" To me it seems a good way to
waste a good dolby A module. But I am willing to be open-minded.... I have dolby A and dolby SR here.
in the studio. A legacy of working on reel-tape.

[Albert H]

Call me old-fashioned, but I always loathed Dolby. I had to line up 24 channels of Dolby A on one occasion, and I swore to never to have anything to do with that junk ever again!

Do you remember when Dolby had that place near to Clapham North? That was back in the days when they were known as "Hire 'n' Fire", since they used to sign up University leavers, and fire them after three months. One friend of mine lasted nearly a year there - probably a record - before he told them exactly what they could do with their job.... I got him a job at De Lane Lea, and he was there (at Wardour Street) for many years.

Just for the sake of amusement, I spent some time lining up my Studer A80, getting the bias exactly right for Ampex 456, and then put it up against an imported (and very expensive) Ampex machine with Dolby A. Mine was about 18dB quieter, had lower distortion, and a flat frequency response. The Ampex - despite spending a lot of time lining it up - was appalling! Even my old - valved - Studer C37 left the Ampex for dead!

If you want some details of Aphex (or rather my implementation of it), I'll be happy to put some circuits up here. My pals at D&R in the Netherlands make their own version of the Aphex box, and still sell it as an outboard module. Their circuits are available on line. They use the LM13700 as a non-linear multiplier stage to generate their harmonic content. It's an interesting approach.

My circuit abused an NE571 (with the filter capacitor about 20 times too small) to generate the harmonics. It's a really good way of generating concordant even harmonic! I selected the band of frequencies to abuse with a tunable state-variable filter with variable Q, so that you could expand the range of frequencies passed to the distortion generator. The funny thing with Aphex - "less is more"! It's easy to turn the ratio knob up too far, since your ears quickly accept quite large amounts of distortion. The trick is to add a very small percentage, and it really makes vocals stand out. The results are similar to your "vocal stressor", but without the need to wreck an expensive Dolby module!