Please identify.

RG> Please identify the Spectrum.
RG>
RG> According to the basic diagram, there are only 67 cases
RG> Board size 300x180mm
RG> Memory 48-separate on Ru6 and Ru5
RG> ROM on two 8k
RG> Tape recorder port on KR1006vi1
RG> The tweeter is located in the lower left corner
RG> The keyboard port is not buffered, as a result
RG> periodic freezing is observed.
RG> On the system connector (or rather, there is no connector, only contact
RG> pads for it) there is a signal ctl-0 and ctl-1, which can be controlled
RG> disable ROM and/or RAM. Supposedly came from
RG> Novosibirsk, but judging by the descriptions I found, this is not a classic
RG> Novosibirsk spec.

Are there PE3 and PT1 chips there?
If there is, then this is a variant called PLM.

There are 3 such ROMs
RT4 - involved in the synthesis of CSROM1, CSROM2 and CSRAM

there are also PT4 and PT5 involved in the video signal and participating in the formation of
INT.

Is this it?
>

--
Sent via the Talk.Ru server - http://www.talk.ru

RG>
RG> There are 3 such ROMs
RG> RT4 - engaged in the synthesis of CSROM1, CSROM2 and CSRAM
RG>
RG> there are also RT4 and PT5 involved in the video signal and participating in
RG> the formation of INT.
RG>
RG> Is this it?

It, it. The variant is called "PLM", developed from a very ancient variant
"NETI". Maybe I'm wrong, but some people called the "PLM" option as an option
"Automation", i.e. option developed at the Institute of Automation on the academic campus.

Of the features, we can note:

1. The designers did not know the architecture of the Spectrum (that there are waiters).

2. When writing to screen occurs, the data goes to both RU5 and RU6. If
the video controller reads RU6 at this moment, i.e. they are zianyat, then the recording is temporarily
postponed using the buffer register.

3. Int comes, in my opinion, at random.

4. There is Turbo, pure 7Mhz. But it was necessary to cheat something to make it work
without failures from the video RAM.

5. Expanding the memory to 128k will be very problematic (IMHO).

---

Subsequently, by 1992-94, PLM was heavily redesigned and a very
good circuit was made for it with an FDD controller, with 128k of memory. In my opinion, they removed the shadow ru6.

But I didn’t see the boards/computers themselves, only the circuit that was lying around somewhere.
Although, I heard that people assembled these computers. The circuit and board were very competent,
it included onlythe best solutions, such as keyboard buffering, proper wiring,
connectors, joysticks, amplifiers, etc.

> without failures with video RAM.
>
> 5. Expanding memory to 128 k will be very problematic (IMHO).
>
> ---
>
> Subsequently, by 1992-94, PLM was heavily redesigned and a
very
> good circuit with an FDD controller was made for it, with 128k memory. In my opinion, they removed the shadow
ru6.
>

RG> This is unlikely. After all, it was 1991, at that time the architecture of
RG> was already well known

I mean the basic architecture, not the re-design. In Novosibirsk, re-design
of boards was often done by completely different people, and not so that there were only
designers, like in St. Petersburg.

RG> No, there RU5 are directly suspended from the brick, and they are accessed
RG> only when they actually access the higher addresses.

Everything is correct, because RU6 and the video screen space are in the lower
half of the memory.

RG> Logically and
RG> are not physically connected in any way with the RU6 line. In turn, for RU6
RG> control signals are generated taking into account the operation of the screen. But on them
RG> there is no register either, but it sticks out in the screenshot.

Recording in ru6 must go through a special buffer register, otherwise there is little sense in
two lines.