Optimod 8100 Lesson

[dannyscott101]

The Prisms can be placed in front of a stock 8100 with great results if:

(1) You put something like a Compellor or other leveler in front of the Prism to keep them in the sweet spot and (2) you let the prisms do all the compression. The 8100 should be running no more than 2-3 db of total gain reduction. Pay particular attention to the “presence band” of the prisms.

Back to the nostalgia part of the thread for a moment... When setting up this combo, I prefer to allow the Compellor to do all the gating.

With an OLD (early) set of Prisms, if you turn the 'gate' control all the way counterclockwise, they will recover to maximum gain (the 'expand' side) instead of the 'return to center' behavior. This way, they can be dialed to any amount of gain reduction desired, controlled by the Compellor feeding them. The "density" of the sound can then controlled by the amount of G/R in the Prisms, and the actual DENSITY control will just adjust the response speed.

This makes the setup operate more like the way an XT-2 behaves, but still maintaining the Prism "sound".

This way, it is also possible to use just a smidge of multiband processing... Like an 8100A with just a little extra Umph! With the normal Prism behavior, you are 'stuck' with a certain type of sound.

The early Texar units will do this. The later Texars and the Gentner units will not, but can be easily modded.

I like this better because some later Prism units seemed to be "Gate Happy"... They would seem to want to duck heavily in certain low level passages, particularly in the mid band, even with a minimum setting on the GATE control. Also when linking two units together, the LEAST desirable behavior seemed to take priority.

I also liked around 10db of G/R in the 8100 following. The Optimod seems to have some degree of 'soft knee' behavior up to around 12dB of G/R. 10dB kept the sound tight enough, but open sounding.

 

[Studio1]

The early Texar units will do this. The later Texars and the Gentner units will not, but can be easily modded.

Where is the delineation between early and late Texars? The pair I have are serial nos 1376 and 1377, and have the phase rotators fitted. If these are 'late' versions, can you detail the mod?

 

[BROADCAST]

If you have the later model where the gate stays on with the gate control fully counter clockwise,you need to jumper (short) R125 on the cx-2 control board.

R125 (220 ohm)was added between the lower end of the gate pot and ground.

 

[radiosaur]

The Prisms can be placed in front of a stock 8100 with great results if:

(1) You put something like a Compellor or other leveler in front of the Prism to keep them in the sweet spot and (2) you let the prisms do all the compression. The 8100 should be running no more than 2-3 db of total gain reduction. Pay particular attention to the “presence band” of the prisms.

Back to the nostalgia part of the thread for a moment... When setting up this combo, I prefer to allow the Compellor to do all the gating.

With an OLD (early) set of Prisms, if you turn the 'gate' control all the way counterclockwise, they will recover to maximum gain (the 'expand' side) instead of the 'return to center' behavior. This way, they can be dialed to any amount of gain reduction desired, controlled by the Compellor feeding them. The "density" of the sound can then controlled by the amount of G/R in the Prisms, and the actual DENSITY control will just adjust the response speed.

This makes the setup operate more like the way an XT-2 behaves, but still maintaining the Prism "sound".

This way, it is also possible to use just a smidge of multiband processing... Like an 8100A with just a little extra Umph! With the normal Prism behavior, you are 'stuck' with a certain type of sound.

The early Texar units will do this. The later Texars and the Gentner units will not, but can be easily modded.

I like this better because some later Prism units seemed to be "Gate Happy"... They would seem to want to duck heavily in certain low level passages, particularly in the mid band, even with a minimum setting on the GATE control. Also when linking two units together, the LEAST desirable behavior seemed to take priority.

I also liked around 10db of G/R in the 8100 following. The Optimod seems to have some degree of 'soft knee' behavior up to around 12dB of G/R. 10dB kept the sound tight enough, but open sounding.

Interesting. I used the Compellor very close to full leveling mode. I let the Prisms do the compression, with minimal gain reduction on the 8100. The limiters were the gems in the 8100. I also never strapped the two prisms together, and I never had the ducking problem. The competition with the 8100 / XT couldn’t get near our loudness without severely mashing the audio.

But, as was mentioned earlier in the thread, once the source material changed with the advent of the horrendous audio the record companies were putting on CD’s, the old adage “garbage in, garbage out” began to assert itself.

 

[fm-engineer]

If you like the specifics of the 8100/XT2 sound texture, you can get closer with a DSP-based Optimod than with other brands because there is still some "XT DNA" in our current DSP-based product line.

Bob Orban

Bob,

I have often wondered why you don't just throw in an 8100/XT2 Preset with the basic settings, just for everyone that wants it...

I have developed two presets that are close to an 8100-XT2 sound. They were A/B'd against the analog version just as they were set in 1995 when removed from a legendary 100kW FM in Virginia. They are formatted for the 8300, 8400, 8500. The analog XT2 version was modified on the bottom end to provide more punchy bass that was in demand in the early 90’s. You can set the bass on the presets very easy to your taste by using the low freq shelf eq. The EQ bass is very touchy since the sections that follow are released very fast. PM me and I will send you the presets to try.

 

[Studio1]

The Prisms can be placed in front of a stock 8100 with great results if:

(1) You put something like a Compellor or other leveler in front of the Prism to keep them in the sweet spot and (2) you let the prisms do all the compression. The 8100 should be running no more than 2-3 db of total gain reduction. Pay particular attention to the “presence band” of the prisms.

When you say "a stock 8100" do you mean an 8100 on its own (without XT2) or an 8100 that hasn't been modified?

 

[LA_Guy]

It was my conclusion (with confirmation from my processing guru pals Gary Blau and Greg Strickland) that the real power in the XT was not its limiters, but rather its clippers. By modifying the limiters to be very slow, but driving relatively deep into the amazingly transparent distortion-cancelling clipping, one could get great openness and loudness at the same time. In my case this involved creating individual threshold trim controls for each band, and removing, or replacing with much larger values, all the release time constant resistors.

Orban's clippers were amazing. I always considered them to be part of the limiter circuit, but of course, they are really after it.

Now for a question for the experts. Since the clippers don't produce much significant distortion, could that design be used to drive streaming audio without the affects that a normal analog clipper has on the lossy compression?

Orban's clippers were 1N914s either straight to ground or off the nose of an opamp . Nothing special there!

 

[LA_Guy]

I always loved the single patent on the 8100 "Distributed crossover with imbedded bass clipper". Bob actually found a way to patent the fact that you can make a two pole filter by putting two one pole filters in series-by putting a set of 1N914 diodes to ground in between them!

 

[rorban]

Orban's clippers were 1N914s either straight to ground or off the nose of an opamp . Nothing special there!

Wrong. Most of the clippers in the 8000 and 8100 were Schottky diodes (HP 5082-2800) biased from rather high voltages (typically around +/- 4 V) to get a much "harder" knee than the "1N914 to ground" circuit would create. The only 1N914-type diodes (actually IN4148s) were in the 8100A's bass clipper and in the XT2 distributed multiband clipper; these diodes were biased from lower voltages to create a softer knee than the clippers in the main audio path.

In addition, the clipper diode biasing supplies were temperature-compensated to keep the clipping threshold constant as temperature varied, which is a "special" design element that several "brand X" processors of the time overlooked.

Bob Orban

 

[rorban]

I always loved the single patent on the 8100 "Distributed crossover with imbedded bass clipper". Bob actually found a way to patent the fact that you can make a two pole filter by putting two one pole filters in series-by putting a set of 1N914 diodes to ground in between them!

The patent in question is U.S. 4,249,042 and there is considerably more to it than that. The implementation in the 8100 exploits the fact that a second-order Linkwitz-Riley crossover has an allpass summation (which means it can contribute one pole to the phase rotator in the signal path) and additionally has poles with a Q of 0.5, which allows the bass signal path to be created as two one-pole filters in series. In the 8100's bass clipper, this means that the clipper is not driven by a filter with a peak gain of more than unity in its frequency response, which minimizes the amount of clipping that occurs in the bass clipper while still allowing the second pole to roll off clipping distortion, again through a filter whose gain never exceeds unity.

The original poster may be unaware of this subtlety (one of the many in the 8100's design) and has evidently had not looked at an 8100 manual before posting, because he would have seen that the diodes (1N4148s) were not "connected to ground" but instead were biased by a bipolar temperature-compensated voltage source to ensure temperature stability of the bass clipping threshold.

Bob Orban

 

[Groove1670]

What kind of results do you get putting a compellor in front of an 8100. Is this a good or bad idea. How does it sound?

 

[rrsounds]

What kind of results do you get putting a compellor in front of an 8100. Is this a good or bad idea. How does it sound?

My opinion would be that you gain nothing, and actually lose something, because the two-band AGC in the 8100 is capable of reducing spectral modulation in ways that are impossible for a single-band compressor such as the Compellor. Never mind that the 8100's AGC is designed as an integral part of the overall system (timing and level-matched), and the Compellor is not.

Using any external processing has to be carefully thought out. It's also my opinion that, at the least, external processing should perform some task that the rest of the system doesn't (or can't) do rather than simply duplicate any particular feature. Don't get me wrong: much of my reputation came from designing external processing boxes! But the negative aspect of putting online an additional piece of gear (an additional point of potential failure) has to be weighed against any perceived advantages.

Kind Regards,
David

 

[Studio1]

It is interesting to see the various comments relating to different ways of driving the 8100.

The station I worked for in the early 90s ran an 8100 with a pair of Prisms in front of it, and the RCF-1 card.

Sadly, the PD kept telling us to 'make it loud' - so the processing was pretty aggressive. [I am still trying to remember if they had an XT2 or not].
Other than that, it sounded great on-air.

 

[wgliradio]

My opinion would be that you gain nothing, and actually lose something, because the two-band AGC in the 8100 is capable of reducing spectral modulation in ways that are impossible for a single-band compressor such as the Compellor. Never mind that the 8100's AGC is designed as an integral part of the overall system (timing and level-matched), and the Compellor is not.

Kind Regards,
David

I do have to add here that the Ariane does wonders in front of an 8100, with the 8100's total G/R not exceeding -10 to -15, it can be very open and still very loud. The Ariane never seemed to fight with the other processors in line with it (set within reason... and you have alot of room with the Ariane).

 

[OKCRadioGuy]

If you use the Compellor, just be careful not to over-do it if you're doing music radio most of the time. I have a 320 ahead of my STL system feeding into a 8100XT/2. Honestly, it sounds excellent. Most of the time we run talk programming, but run some music. It sounds great for both types of programming. I have some Omnia One's also in some of our extended markets past the metro (non-comm). The seem a bit harsh to me on voice. I really like the sound of the analog Orban stuff better for what we are doing most of the time. The Compellors really help smooth out the TERRIBLE levels from the PRSS bird (NPR and other public networks). Without the Compellor, we'd have a real mess on the air.

 

[radiosaur]

The Prisms can be placed in front of a stock 8100 with great results if:

(1) You put something like a Compellor or other leveler in front of the Prism to keep them in the sweet spot and (2) you let the prisms do all the compression. The 8100 should be running no more than 2-3 db of total gain reduction. Pay particular attention to the “presence band” of the prisms.

When you say "a stock 8100" do you mean an 8100 on its own (without XT2) or an 8100 that hasn't been modified?

Sorry, late response but the answer is an unmodified 8100 without the XT. I also tried an RCF-1.

It absolutely killed the loudness of the chain.

 

[Groove1670]

Going to add a XT2 to our 8100A/1. In need to modify the Power supply resistors or the XT will not work properly. Is it better to piggyback the resistors R103 & R104 with 1.8 ohm resistors (as suggested in the manual or replace them with the proper value (I think it is 62 ohms?). Has anyone done this?

 

[Studio1]

The Prisms can be placed in front of a stock 8100 with great results if:

(1) You put something like a Compellor or other leveler in front of the Prism to keep them in the sweet spot and (2) you let the prisms do all the compression. The 8100 should be running no more than 2-3 db of total gain reduction. Pay particular attention to the “presence band” of the prisms.

When you say "a stock 8100" do you mean an 8100 on its own (without XT2) or an 8100 that hasn't been modified?

Sorry, late response but the answer is an unmodified 8100 without the XT. I also tried an RCF-1.

It absolutely killed the loudness of the chain.

Damn, I think I've lost the picture here. Are you saying that you shouldn't use the RCF-1 when you're running Prisms?

 

[TomT]

(Up one post)--

It's a chore to get to that power supply card, but if you want to take the time, yes, go ahead and replace the resistors with the correct value. While you are at it, though, take time to replace the two electrolytic caps on the same board. Most 8100's are at least 15 years old or older, so is best to replace these caps as they are getting pretty dried out.

You need to pull the two regulating transistors (under the white plastic covers on the back) out of their sockets. Next remove the entire back panel (lots of screws--get a coffee cup and through all the hardware into it). Then, using needle nose, you can gently lift the board off the white plastic stand-offs.

Solder wick will be helpful to loosen the parts up--Radio Shack's type works pretty good.

 

[Groove1670]

Thanks Tom,

The caps have been replaced. I just got the XT2. Hence the modification. I'm just thinkin' it would be easier to piggyback the 1.8 ohm resistors. How does the xt2 malfunction with the 91 ohm resistors currently in place.